MPS IIIA · Gene therapy · Regulatory milestone
Ultragenyx: EMA validates the marketing authorisation application for rebisufligene etisparvovec (FAYUVI / UX111) in MPS IIIA (2 Oct 2026).
The EMA has validated Ultragenyx’s marketing authorisation application for the single-dose intravenous AAV9 SGSH
gene therapy already FDA-approved (September 17, 2026, as FAYUVI) for neurologic MPS IIIA in pediatric patients
with preserved neurodevelopment. Formal EU scientific review now starts. This is not EU approval.
Caveat: Validation is not marketing authorisation, a safety or efficacy determination, or an access guarantee. PRIME and Orphan designations do not assure approval. Distinct from the earlier U.S. FDA approval already in the archive.
Ultragenyx via GlobeNewswire →
Beta-thalassemia · CRISPR · Routine care
Charité administers Germany’s first routine-care exagamglogene autotemcel (Casgevy) for transfusion-dependent beta-thalassemia (1 Oct 2026).
A 19-year-old with severe beta-thalassemia received autologous CRISPR-Cas9-edited hematopoietic stem cells
(gamma-globin reactivation) at Charité, the first qualified German center. Dosed on May 28, 2026, the patient
was transfusion-independent by about four months, with fetal hemoglobin about 85% of total hemoglobin.
Caveat: Conditional EU approval; 15-year monitoring is required. Chemotherapy conditioning carries mucositis, a likely risk of infertility, and possible liver injury. This is a single-center access story, not new pivotal efficacy data. Long-term safety is still accruing.
EurekAlert / Charité →
Base editing · Research embryos · Not clinical
Adenine base editing rewrites PCSK9 in human research embryos without Cas9 double-strand breaks (Nature; Scienmag, 2 Oct 2026).
Protein-delivered ABE8e-V106W rewrote all PCSK9 alleles and supported development to the blastocyst stage, with
no target-site insertions or deletions detected, cleaner than Cas9 double-strand-break editing in embryos. The
same study edited HBG. The authors report rare on-target chromosome breakage, mosaic bystander and off-target
edits, and developmental arrest when the editor was delivered as mRNA because of guide-independent deaminase
activity. They say these risks currently preclude clinical use in reproduction.
Caveat: Research embryos only. Not a therapy. Not a green light for heritable editing.
Scienmag →
Nature (DOI) →
Hematopoietic stem cells · Imaging AI · Preclinical
ChromAgeNet: deep learning predicts hematopoietic stem-cell aging from 3D chromatin images (Aging Cell, 28 Sep 2026).
A convolutional network trained on DAPI-stained three-dimensional images of mouse hematopoietic stem-cell nuclei
distinguishes young from aged cells with an area under the curve of about 0.77 (reported in coverage as about a
77% chance of a correct call), using chromatin entropy and peripheral heterochromatin. It was piloted as a screen
for epigenetic drugs that shift aged nuclei toward a younger-looking chromatin pattern.
Caveat: Mouse cells. An image phenotype is not functional rejuvenation. This is a research tool complementary to epigenetic clocks, not a clinical diagnostic.
IDIBELL →
EurekAlert →
Aging Cell (DOI) →
Senescence · Proteomics · Observational
SASP Score: a deep-learning blood measure of senescent-cell secretory burden (Lifespan.io, 28 Sep 2026).
A guided autoencoder plus transformer, trained on about 38 senescence-associated secretory proteins from about
51,000 UK Biobank participants (50,997 in the proteomics set), produces a blood score of overall senescent-cell
secretory burden. A higher score was associated with about 1.4 times the all-cause mortality and with stronger
links to chronic kidney disease, dementia, and stroke after covariate adjustment. In the MEDEX exercise cohort,
scores stayed roughly flat in the exercise group while they rose in the non-exercise group.
Caveat: A supplementary senescence-burden metric, not a general aging clock. It cannot specify which disease. Observational associations are not a proven intervention surrogate.
Lifespan.io →
Preprint →
