Immune Therapy Engineered Inside the Body Eases Multiple Sclerosis

Posted on 7 September, 2026 by Mehrdad Fathi

For years, CAR-T-cell therapy has operated on a straightforward but labour-intensive logic: extract T cells from a patient, engineer them in a laboratory to target a specific molecule, and reinfuse them. It works. It has transformed outcomes in blood cancers. But it is slow, expensive, and operationally demanding in ways that limit who can access it and where.

A clinical trial published this week in The New England Journal of Medicine tests a different premise: what if the genetic reprogramming happened inside the body?

The In Vivo Approach

The trial enrolled 16 participants with multiple sclerosis and related autoimmune conditions — diseases in which immune cells attack healthy tissue. Rather than extracting and editing T cells ex vivo, the team administered a single intravenous injection of a modified lentivirus, developed by Shenzhen Genocury Biotech, carrying genetic instructions for chimaeric antigen receptors (CARs). Once inside the body, the virus transferred those instructions to circulating T cells, which then began expressing CARs capable of targeting autoreactive B cells — the immune cells producing the antibodies driving tissue damage.

Over the following months, participants generated increasing numbers of CAR T cells. B cells depleted. Autoantibody levels fell. And critically, when B cells repopulated, they did so without producing the disease-associated autoantibodies — a finding the researchers describe as evidence of immune system reset.

Clinically, participants with multiple sclerosis showed improvements in motor and cognitive function and reductions in fatigue. Those with muscle-affecting autoimmune conditions showed improved strength scores and decreased inflammation markers.

Why This Matters

The significance operates on two levels.

First, there is the biology. Conventional ex vivo CAR-T therapies for autoimmune disease are already generating compelling signals in early trials — but they carry the overhead of complex manufacturing pipelines. An in vivo approach, if it holds up, bypasses that infrastructuy the overhead of complex manufacturing pipelines. An in vivo approach, if it holds up, bypasses that infrastructuy in ways that matter for access.

Second, there is the mechanistic implication. The apparent immune reset — replacement B cells that do not produce pathogenic autoantibodies — suggests the treatment may be doing something more durable than suppression. It may be reoantibodies — suggests the treatment may be doing something more durable than suppression. It may be rems.

David Simon at Charité — University Medicine Berlin describes the results as “a very exciting proof-of-concept study.” Bing Du at East China Normal University calls them “good and comparable with those for ex vivo CAR-T-cell therapy.” Dai-Shi Tian, one of the trial’s lead investigators, is more measured: “The responses are promising signals, but they are not yet definitive evidence of efficacy or permanent restoration of immune tolerance.”

What Remains Open

The safety profile was manageable: most participants experienced mild, transient inflammatory responses; three showed temporarily reduced white blood cell counts that subsequently recovered. But the trial was small, lacked a control group, and followed participants for roughly six months.

The larger question is longer-term safety. Because lentiviral vectors integrate genetic material into the host genome, there is a real, if low, risk of insertional mutagenesis — the inserted sequence disrupting a gene in ways that could eventually promote tumour development. Participants will be monitored for several years for cancer, infections, and disease recurrence. Du suggests a ten-year minimum follow-up is warranted before clinical confidence is established.

The team intends to move toward larger, disease-specific studies. The path from proof-of-concept to a scalable therapeutic is long, and Simon’s framing is probably the right one: this “could be a gamechanger” — conditional on confirmation in properly powered, controlled trials with extended follow-up.

For now, the data make a credible case that in vivo CAR-T-cell therapy for autoimmune disease is feasible, that immune reprogramming inside the body produces biologically meaningful responses, and that the approach is worth pursuing at scale.

Source: Cheng, Y.-H. et al. N. Engl. J. Med. 395, 926–929 (2026). Reported in Nature (2026), doi: https://doi.org/10.1038/d41586-026-02765-1

References:

1. Cheng, Y.-H. et al. N. Engl. J. Med. 395, 926–929 (2026).

2. Xu, J. et al. Lancet 406, 228–231 (2025).