Primary data from a phase 1/2 clinical trial of an investigational dual-vector gene therapy for GM2 gangliosidosis, published in Nature Medicine,1 suggest that the approach resulted in measurable enzymatic activity, substrate reduction, and encouraging neurologic outcomes in patients with infantile Tay–Sachs disease (TSD) and Sandhoff disease (SD).1,2 The therapy, delivered directly to the thalamus and cerebrospinal fluid (CSF), demonstrated dose-responsive target engagement and was generally well tolerated.
These findings mark the first clinical evidence that intraparenchymal gene therapy targeting deep brain structures may be a viable therapeutic strategy for infantile GM2 gangliosidosis, a rare and fatal lysosomal storage disorder that currently lacks approved disease-modifying treatments.
Importantly, patients experienced positive effects on their ability to remain on oral feeding regimens—historicaly, more than 50% of patients with GM2 gangliosidosis need to be fed intravenously—and multiple partial effects related to seizure onset.
"This is encouraging because eating by mouth in an important quality of life outcome for the families of these children,” study investigator Heather Gray-Edwards, DVM, PhD, assistant professor of genetic & cellular medicine at UMass Chan, and a member of the Horae Gene Therapy Center, said in a statement.2 “Nonetheless, the partial effects of the therapy indicate the need for continued improvements of the gene therapy."
Enzymatic and Substrate Activity Improvements With AAV Gene Therapy
In total, 9 patients were treated (TSD, n = 7; SD, n = 2), including 6 with infantile and 3 with juvenile forms of GM2 gangliosidosis. The dual-vector approach used separate recombinant adeno-associated virus serotype rh8 (rAAVrh8) constructs to deliver HEXA and HEXB genes, administered via bilateral thalamic infusion and both intra-cisterna magna and lumbar intrathecal injections.
In treated infantile patients, dose-dependent increases in CSF β-hexosaminidase A (HEXA) activity were observed, peaking at 0.59 nmol/h/mL—or 13% of the normal mean—by 12 weeks post treatment. Total serum HEX activity exceeded 40 nmol/h/mL in high-dose recipients, surpassing the lower limit of the reference range.
Biochemical substrate reduction was also observed, with CSF C20:0 GM2 ganglioside levels declining in 4 of 6 infantile patients by up to 52.5% from baseline. These changes suggest that the expressed enzyme reached relevant compartments and was functionally active in degrading stored ganglioside.
Data From the 2023 ASGCT Meeting
Presented at the 2023 American Society of Gene and Cell Therapy (ASGCT) Annual Meeting by author Terence R. Flotte, MD, the phase 1 trial data of the 2-vector rAAVrh8 gene therapy delivered via bilateral intrathalamic and CSF injection was well-tolerated and showed partial stabilization of Tay-Sachs and Sandhoff disease in pediatric patients.
All patients showed decreased CSF GM2 levels and increased HEXA activity. Notable findings included preserved oral feeding, delayed seizure onset, and radiologic signs of brain development in infants. Due to dystonia in juveniles, future studies will focus on infants and explore a bicistronic vector.
READ MORE: Dual Injections of 2-Vector Gene Therapy May Stabilize Disease in GM2 Gangliosidosis
Additionally, structural imaging and clinical outcomes aligned with these biochemical improvements. Diffusion tensor imaging (DTI) showed increased fiber tract numbers and fractional anisotropy in white matter regions including the corpus callosum and lentiform nucleus—changes that persisted beyond 12 weeks.
Volumetric MRI scans indicated slower brain volume growth relative to untreated historical controls, with reduced storage burden inferred from lower volumetric expansion. Patients with lower brain volumes showed correspondingly milder clinical severity scores.
Clinically, half of the infantile patients maintained oral feeding ability past 25 months—well beyond the typical window when gastrostomy is required in untreated patients. One high-dose patient retained full oral feeding through 27 months of age.
Delayed seizure onset and improved seizure control were also noted. Two high-dose patients either remained seizure-free or achieved control with a single anti-epileptic drug during the 24-week study period.
Global Clinical Impression (GCI) scores indicated early neurologic stabilization or improvement in several patients, particularly those who received the higher vector doses. One high-dose recipient (Patient 011) maintained neurologic gains through week 24, correlating with the highest post-treatment enzyme activity levels in the cohort.
AAV Gene Therapy Safety Profile and Juvenile Disease Challenges
Across the trial, 171 adverse events (AEs) were reported, 15 of which were considered possibly or definitely related to the gene therapy. Most treatment-related AEs included transient liver enzyme elevations and immune responses to the AAVrh8 capsid. These events responded to corticosteroids and did not result in treatment discontinuation.
Sixteen AEs were associated with the surgical procedure or anesthesia. These were managed with protocol modifications, including extended suture retention to prevent wound dehiscence.
No treatment-related deaths occurred. Two deaths—1 from disease progression and another from Clostridioides difficile infection—occurred but were not attributed to the therapy.
Notably, all 3 juvenile patients exhibited either worsening dystonia or neurologic decline following treatment. While the cause has been noted as unclear, the investigators elected to exclude juvenile patients from further enrollment and analysis. Imaging findings suggested possible thalamic inflammation, but no clear causal link was established.
Expert Perspective and Future Direction for HEXA / HEXB Gene Therapy
“Biochemically, it worked,” Gray-Edwards said in a statement.2 “We were able to induce production of the appropriate enzyme and that enzyme was functional. Although we didn’t achieve therapeutic levels, our thalamic injections ended up being safe in patients and the transgene vectors work. That’s an important step.”
The investigational gene therapy was developed in collaboration with UMass Chan Medical School and supported by several patient advocacy groups, including the National Tay-Sachs & Allied Diseases Association, Cure Tay-Sachs Foundation, and the Blu Genes Foundation.
“This research is an example of the important work that our faculty is doing in our Translational Institute for Molecular Therapeutics,” Flotte, the Elisabeth Chair for the Dean of Medicine, executive deputy chancellor, provost and dean of the TH Chan School of Medicine, said in a statement.2 “Driven by Dr. Gray-Edwards and Dr. Sena-Esteves, the institute leverages our extensive experience in researching and developing gene therapies for early-stage clinical trials so we can play a key role in moving therapies for rare diseases forward. It is so important to patients and families like those who participated in these trials that our research is providing hope for some help for this devastating disease.”
Given the existing limitations of dual-vector dosing and delivery, the research team is now pursuing a bicistronic vector platform to simplify administration and potentially expand clinical applications. A next-generation construct coexpressing both HEXA and HEXB genes in a single AAV vector is currently in development.
This trial represents a major milestone in the treatment of infantile GM2 gangliosidosis, laying the groundwork for future trials aimed at improving durability, delivery efficiency, and accessibility of gene-based approaches for lysosomal storage disorders.
REFERENCES
1. Eichler F, Cataltepe OI, Daci R, et al. Dual-vector rAAVrh8 gene therapy for GM2 gangliosidosis: a phase 1/2 trial. Nat Med. ePub 2025. doi:10.1038/s41591-025-03822-4
2. Phase I/II clinical study of gene therapy for GM2 gangliosidosis, including Tay-Sachs and Sandhoff diseases, shows encouraging results. News release. UMass Chan Medical School. August 18, 2025. Accessed August 29, 2025. https://www.eurekalert.org/news-releases/1095058
