Dual-Vector Editing Maintains Year-Long Function for DMD

Duchenne muscular dystrophy results from variants that disrupt the reading frame of the X-linked DMD gene, preventing production of functional dystrophin. Without dystrophin, the connection between the muscle-cell cytoskeleton and extracellular matrix becomes unstable.

GEN6050X – developed by Chinese-based GenAssist Therapeutics – is intended for patients whose variants are amenable to skipping exon 50 in DMD. Removing this exon from the mature transcript can restore the reading frame and produce a shorter dystrophin similar to forms associated with the generally milder Becker muscular dystrophy phenotype.

The FDA cleared GenAssist’s Investigational New Drug application for GEN6050X on 6 March 2025, allowing the company to begin clinical testing in the United States. This was followed by Orphan Drug Designation in June 2025 and Rare Pediatric Disease Designation later that month. IND clearance and these designations do not constitute approval of GEN6050X for clinical use or confirmation that it is safe or effective.

GEN6050X combines CRISPR-guided cytosine base editing with gene addition. A single intravenous infusion delivers two single-stranded AAV9 vectors. The first, ss.AAV9.oTAM, encodes a Targeted-AID-mediated mutagenesis cytosine base editor under a muscle-specific promoter.

GenAssist describes oTAM as an RNA-editing-free editor, although its complete architecture and promoter sequence have not been disclosed. The second vector, ss.AAV9.hE50-sgRNA, carries three copies of a guide RNA targeting the 5′ splice site between DMD exon 50 and intron 50.

Both vectors must reach the same cell: one supplies the editor and the other its sequence-specific guides. The editor is designed to convert a cytosine at the splice site without generating a double-strand DNA break. Disrupting this site causes the splicing machinery to omit exon 50 from some DMD transcripts, restoring an open reading frame in eligible genotypes.

Unlike antisense exon skipping, which acts transiently on RNA and requires repeated treatment, successful DNA editing could provide a lasting source of exon-skipped transcripts. Public materials do not disclose the ratio between the vectors, the guide-expression promoter or editing frequencies in all three patients.

The guide-carrying vector also contains a human ACTG1 expression cassette. This is gene addition rather than editing of the endogenous ACTG1 locus: transduced cells receive an extra template encoding cytoplasmic γ-actin. A GenAssist poster states that the cassette is controlled by a "HAS" promoter but does not explain the abbreviation or provide its sequence.

The company proposes that additional γ-actin could interact with newly produced dystrophin and accelerate reconstruction of costameres and the dystrophin-associated glycoprotein complex. The clinical contribution of this component cannot be separated from dystrophin restoration because both interventions are administered together.

Early signals suggest benefit but need confirmation

In induced pluripotent stem cell-derived myotubes carrying a deletion of DMD exons 51–53, quantitative PCR detected exon 50 skipping and exogenous ACTG1 transcripts seven days after exposure. At multiplicities of infection of 5 × 10⁵ and 1 × 10⁶, mean exon-skipping proportions were 47.42% and 59.10%, respectively. Corresponding ACTG1 transcript measurements averaged 2.6 × 10³ and 3.2 × 10³ copies per nanogram of total RNA. These product-potency experiments demonstrate activity of both components in cultured cells but cannot predict clinical efficacy.

The investigator-initiated trial enrolled three ambulatory boys aged 6.5–10 years at Peking Union Medical College Hospital. Each received GEN6050X once at 5 × 10¹³ vector genomes/kg. In an earlier six-month update on the first two participants, GenAssist reported modest improvements in North Star Ambulatory Assessment and Performance of Upper Limb 2.0 scores.

The first participant reported improvement of more than 100 metres (27%) in the six-minute walk test, which exceeds the approximately 30-metre change generally regarded as clinically meaningful in DMD. The second patient reported a 2% decrease in the same test.

Hand-held dynamometry indicated that elbow and knee strength in the first participant varied between a 26% decline and an 18% increase, while the second patient showed increases in upper- and lower-limb strength, including improvements of 11–50% in elbow strength.

The same update reported 1.6% DNA editing and 2.2% exon 50 skipping in a muscle biopsy from the second participant, with dystrophin abundance rising to 2.9 times its baseline level. Editing, exon skipping and dystrophin were not detected in the first participant’s biopsy, which the company attributed to high fat content and resulting analytical limitations. These biomarker data are based on single samples and do not establish a relationship between editing frequency, dystrophin restoration and motor performance.

At the subsequent one-year assessment, motor function remained stable or showed modest improvement overall. Mean left ventricular ejection fraction increased by 5 percentage points, although the standard deviation of 16 indicates substantial variation. Mean relative forced vital capacity increased by 12%, while peak expiratory flow increased by 25%. The release did not provide individual trajectories, confidence intervals or updated patient-level editing and dystrophin results. Age-related variation and the absence of a comparator further limit interpretation.

Transient serious adverse events occurred during inpatient monitoring and reportedly resolved without sequelae. Their nature and relationship to treatment were not disclosed. No new clinically significant treatment-related symptoms or laboratory abnormalities were reported during longer follow-up. Earlier testing found no detectable anti-Cas9 IgG antibodies or T-cell responses against AAV9, the transgene or dystrophin in the first two participants, but three patients are insufficient to identify uncommon or delayed toxicities.

The findings therefore provide preliminary observations of biological activity and safety rather than comparative evidence of benefit. Patient-level functional results, muscle-biopsy measurements and larger controlled cohorts will be required to determine whether splice-site editing, ACTG1 delivery or their combination produces durable clinical effects.

This article is based on press releases from GenAssist Therapeutics on 24 July 2026 and 17 May 2025.