Two Roads to the Same Goal: AAV Gene Therapy vs. Ex Vivo Lentiviral Gene Therapy for GM1

A plain-language guide for families, what these approaches are, how they differ, and what a new research paper adds to the picture.

Medically reviewed by Simon A. Jones, MBChB, BSc, MRCPCH.

Starting with the problem

GM1 gangliosidosis happens because a gene called GLB1 doesn’t work properly. That gene is the instruction manual for making an enzyme, beta-galactosidase, whose job is to break down a fatty molecule (GM1 ganglioside) inside the cell’s recycling centers, the lysosomes. Without enough working enzyme, that molecule builds up and slowly damages the brain and nervous system.[1]

Every gene therapy for GM1 is trying to solve the same root problem: get working copies of the GLB1 instructions, or the enzyme they produce, into the cells that need them. Where the approaches differ is how they deliver the fix and where the correction happens. Below are the two main strategies, plus what a newly published study adds.

---

Approach 1: AAV gene therapy, “fix the cells inside the body”

The idea in one sentence: Use a harmless virus as a delivery truck to carry a working copy of the GLB1 gene directly into the patient’s own cells, so those cells start making the missing enzyme themselves.

How it works:

• Scientists take a virus called AAV (adeno-associated virus), a virus that does not cause disease in people, and empty out its own genetic material.
• In its place, they load a working copy of the GLB1 gene.
• This engineered virus is infused into the patient, either into a vein (intravenous, so it travels throughout the body) or directly into the fluid around the brain and spinal cord (to concentrate the dose in the nervous system).
• The virus enters cells and delivers the GLB1 gene. Those cells then read the new instructions and begin producing beta-galactosidase. Neighboring cells can also pick up enzyme that gets released, a helpful bonus called “cross-correction.”

Key features for families to understand:

• This is done “in vivo,” Latin for “in the living body.” The correction happens inside the patient; nothing is removed.
• It is typically a one-time infusion.
• The AAV usually delivers the gene as a separate instruction sheet that sits alongside the cell’s own DNA rather than being stitched into it. This works well in cells that don’t divide much, like neurons, but the effect can dilute in cell types that keep dividing.
• The immune system can react. The body naturally fights off viruses, and in GM1 it may also react to beta-galactosidase itself, because the enzyme is new to it. To manage this, doctors typically give medicines that suppress the immune system during treatment and for a while afterward.

Where it stands: AAV gene therapy is the approach furthest along in human GM1 trials. Current examples include an intravenous AAV9 gene therapy run by the NIH (NCT03952637)[2] and PBGM01, delivered into the brain, run by Gemma Biotherapeutics (NCT04713475)[3].

What the human data show so far: The first 3-year results from the NIH intravenous AAV9-GLB1 trial were published in The New England Journal of Medicine in 2026[4]. Nine children with Type II GM1 (ages roughly 2.7–7.3 years) received a single intravenous infusion at one of two dose levels. Three years later:

• Biological signs that the therapy is working: treated children showed increases in beta-galactosidase enzyme activity and reductions in GM1 ganglioside, and, strikingly, brain imaging showed gains in neuronal fiber tracts in the younger (late-infantile) children, in contrast to the steady decline seen in untreated GM1.
• Safety: a single IV infusion was generally tolerated over 3 years; the main issues were severe vomiting (one case serious enough to require hospitalization for IV fluids, the only serious adverse event attributed to the gene therapy) and temporary elevations in liver blood tests. There were no treatment-related deaths reported in the 3-year analysis.
• The honest limits: on standardized developmental scales (Vineland-3), most children’s scores did not change beyond measurement error, and some skill areas (such as fine motor and receptive communication) still declined. In other words, the therapy showed clear biological effects and encouraging brain-imaging signals, but this early trial did not demonstrate broad reversal of the disease’s clinical course. It is an open-label, dose-finding study in a small number of children, an important first step.

---

Approach 2: Ex vivo lentiviral gene therapy, “fix the cells outside the body, then give them back”

The idea in one sentence: Remove some of the patient’s own blood-forming stem cells, correct them in the lab, and return them, so the patient’s body grows a permanent, self-renewing supply of enzyme-producing cells.

How it works:

• Doctors collect the patient’s hematopoietic stem cells (HSCs), the master cells in bone marrow that produce all blood and immune cells for life.
• In the lab, these cells are corrected using a different kind of virus, a lentivirus, carrying a working GLB1 gene. A key difference: the lentivirus integrates the new gene permanently into the cell’s own DNA, so every future cell that descends from that stem cell also carries the fix.
• The corrected stem cells are infused back into the patient (an autologous transplant, “autologous” means the cells are the patient’s own, so there’s no risk of donor rejection).
• These cells engraft in the bone marrow and become a lifelong factory. Crucially, some of their descendants are immune cells that can travel into the brain and become microglia-like cells, delivering enzyme to the nervous system from the inside.

Key features for families to understand:

• This is done “ex vivo,” Latin for “outside the living body.” The genetic correction happens in the lab, in the patient’s own cells, before they’re returned.
• Because the fix is built into stem cells that renew themselves for life, it is designed to be durable and self-sustaining.
• It usually requires conditioning (a type of chemotherapy, similar to what precedes a bone-marrow transplant) to make room for the corrected cells to take hold, a more involved process than a single infusion.
• This is the same general strategy already approved and proven in other lysosomal and related diseases, most notably metachromatic leukodystrophy (MLD), where ex vivo lentiviral HSC gene therapy has been transformative. The research groups behind the GM1 work have deep roots in that success.

---

What the Poletti publication adds

In December 2025, a team led by Valentina Poletti and Alessandra Biffi published a study online in the journal Molecular Therapy titled “Lentiviral hematopoietic stem cell gene therapy ameliorates GM1-gangliosidosis in mice.”[5]

What they did, in plain terms:

• They built an ex vivo lentiviral HSC gene therapy for GM1 (Approach 2 above) using a “last-generation” lentiviral vector deliberately designed to be translatable to human patients. They tested two versions: one carrying a human codon-optimized GLB1 gene and one carrying the mouse version of the enzyme.
• First, in patient-derived cells in the lab, the lentivirally corrected cells produced and secreted working beta-galactosidase, and those corrected cells showed complete clearance of their lysosomal storage. Separately, when untreated GM1 cells were bathed in the secreted enzyme, they took it up and regained beta-galactosidase activity inside the cell, uptake that the authors showed happens through the mannose-6-phosphate receptor, the cell’s normal enzyme-import doorway. Together these results support the principle of “cross-correction,” in which corrected cells can supply enzyme to their neighbors.
• Then they tested it in a mouse model of GM1 gangliosidosis, mice that make no beta-galactosidase and develop tremors, balance and gait problems, and ataxia from about 12–14 weeks of age, with shortened survival (about 40–45 weeks).
• Corrected blood stem cells were transplanted after busulfan conditioning (a standard preparative treatment, as in a bone-marrow transplant). Crucially, they compared a standard intravenous infusion of the corrected cells against an innovative combined delivery, giving the cells both intravenously and directly into the fluid spaces of the brain (intracerebroventricular), to get more corrected cells into the nervous system.

What they found:

• The corrected stem cells engrafted well (donor cells made up a median of over 80% of the bone marrow), and the therapy restored beta-galactosidase activity in the blood, bone marrow, and, importantly, the brain.
• It reduced the toxic storage buildup in the brain and improved the mice’s motor performance (balance beam, rotarod, and wire-hang tests), keeping them at a level between healthy mice and untreated GM1 mice, and delayed their decline by roughly 10 weeks.
• It produced a significant increase in survival. Notably, the group that received the human enzyme by the combined delivery route survived as long as healthy (wild-type) mice.
• Consistent with the known biology, the correction in the brain is delivered by donor-derived cells that migrate in and become microglia-like cells, resident immune cells of the brain, which then supply enzyme to their neighbors.

The honest context every family should keep in mind: This is a preclinical mouse study, not a human trial. The authors were also candid about its limits, the correction of the neurological phenotype was only partial, and more work is needed. It is an important proof-of-concept: it shows that the ex vivo lentiviral strategy, already successful in other diseases, can meaningfully help against GM1 in an animal model, and that adding direct brain delivery may boost the benefit, but it is an earlier step on the path than the AAV therapies currently being tested in people. There is no ex vivo lentiviral GM1 therapy available to patients yet; this work builds the scientific case for developing one. It is also clear from other lentiviral treatments that this type of therapy is better at preventing disease than reversing it. This makes clinical trial design complicated.

---

Side by side

	AAV gene therapy	Ex vivo lentiviral HSC gene therapy (the Poletti/Biffi approach)
Where correction happens	Inside the body (in vivo)	In the lab, then cells returned (ex vivo)
Delivery vehicle	AAV virus (non-integrating)	Lentivirus (integrates into the cell’s DNA)
What’s treated	The patient’s own cells, directly	The patient’s own blood stem cells
How it’s given	Single infusion, IV or into brain/spinal fluid	Cells collected, corrected, transplanted back; usually needs conditioning
How it reaches the brain	Virus crosses into or is placed near the nervous system; enzyme cross-corrects neighbors	Corrected immune cells migrate into the brain as microglia-like cells and supply enzyme there
Durability	Long-lasting in non-dividing cells; can dilute in dividing ones	Designed to be lifelong and self-renewing
Complexity for the patient	Lower, one procedure	Higher, cell collection, conditioning, transplant
Stage for GM1 (2026)	In human clinical trials	Preclinical (mouse); proven in other diseases like MLD

---

So which is “better”?

There is no single winner, and that’s actually good news for families. The two approaches have different strengths, and both are being pursued precisely because GM1 is a whole-body and whole-brain disease that may need more than one tool:

• AAV offers a simpler, one-time treatment and is already in clinical trials for GM1. Its challenge is delivering enough gene, durably, to every affected tissue.
• Ex vivo lentiviral offers a durable, self-renewing supply of enzyme and a natural route into the brain through the immune system, with a strong track record in related diseases. Its challenges are the more involved procedure and the fact that, for GM1 specifically, it is still at the research stage.

The Poletti publication matters because it widens the field. Rather than betting everything on one strategy, researchers are building a portfolio of approaches, AAV, ex vivo lentiviral, enzyme replacement, and substrate reduction, increasing the odds that one, or a combination, will reach patients.

---

The bottom line for families

• AAV gene therapy = a virus delivers a working GLB1 gene into your child’s cells, inside the body, in a single infusion. This is the approach currently in human trials for GM1.
• Ex vivo lentiviral gene therapy = your child’s own blood stem cells are corrected in the lab and given back, becoming a lifelong enzyme factory that also seeds the brain. The recent Poletti/Biffi paper shows this works in mice with GM1 and lays the groundwork for a possible future human trial.
• Both aim at the same target, restoring beta-galactosidase, by different, complementary routes. Having multiple serious approaches in development is one of the most hopeful signs in GM1 research today.

---

Frequently asked questions

Learn more: About GM1 Gangliosidosis · AAV Gene Therapy for GM1 · Enzyme Replacement Therapy for GM1 · GM1 Clinical Trials Guide · History of GM1 Gangliosidosis · Join the GM1 Matrix

---

References

1. Regier DS, Tifft CJ, Rothermel CE. GLB1-related disorders. In: GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 2013 Oct 17 [updated 2021 Apr 22]. https://www.ncbi.nlm.nih.gov/books/NBK164500/ Accessed August 14, 2026.
2. National Human Genome Research Institute. A phase 1/2 study of intravenous gene transfer with an AAV9 vector expressing human beta-galactosidase in type I and type II GM1 gangliosidosis. ClinicalTrials.gov identifier NCT03952637. https://clinicaltrials.gov/study/NCT03952637 Recruiting; status checked August 14, 2026.
3. Gemma Biotherapeutics. Phase 1/2 open-label, multicenter study to assess the safety, tolerability and efficacy of a single dose of PBGM01 delivered into the cisterna magna of pediatric type 1 (early onset) and type 2a (late onset) infantile GM1 gangliosidosis. ClinicalTrials.gov identifier NCT04713475. https://clinicaltrials.gov/study/NCT04713475 Active, not recruiting; status checked August 14, 2026.
4. Lewis CJ, D’Souza P, Johnston JM, et al. AAV9 gene therapy in type II GM1 gangliosidosis: a phase 1-2 trial. N Engl J Med. 2026;394(12):1184-1194. doi:10.1056/NEJMoa2510935. PMID 41665410.
5. Bucciarelli L, Fabris C, Accardo M, Biffi A, Poletti V. Lentiviral hematopoietic stem cell gene therapy ameliorates GM1-gangliosidosis in mice. Mol Ther. 2026;34(3):1854-1866. doi:10.1016/j.ymthe.2025.12.007. PMID 41376163.

Last updated: August 21, 2026.

This article is for educational purposes and does not constitute medical advice. Diagnostic and treatment decisions should be made with a qualified healthcare professional. The lentiviral study described here was conducted in mice; no ex vivo lentiviral gene therapy for GM1 is currently available to patients. Study details are drawn from the published papers (which remain under their publishers’ copyright) and were verified against those publications, GeneReviews, and ClinicalTrials.gov (2026).