Coverage window: September 14-20, 2026. Published September 22.
A retinal cell graft is still distinguishable ten years after transplantation. That is the week's most consequential observation, even though it comes from one patient and cannot establish how much vision the procedure preserved. Long follow-up gives regenerative medicine something that short experiments cannot: a view of what remains.
Elsewhere, the field is improving how cells are qualified, programmed, and measured. This issue follows an autologous Parkinson's manufacturing strategy, self-amplifying RNA for iPSC experiments, and a retinal-vesicle study whose proposed mechanism can now be tested more sharply. Two MS papers and an Alzheimer’s preprint show why biological activity, model discrimination, and clinical benefit need separate judgments.
1. A first-in-human iPSC retinal graft reaches ten-year follow-up
Primary source: Ophthalmology Retina article record and abstract
A September 16 report revisits the patient who received an autologous iPSC-derived retinal pigment epithelium sheet in 2014. She had neovascular age-related macular degeneration and was 88 at the ten-year follow-up. Multimodal imaging continued to distinguish the pigmented graft area, with preservation of the overlying outer nuclear layer and underlying choroidal structure.
The imaging is more informative than a photograph of pigment alone. Polarization-sensitive OCT detected a persistent melanin-associated signal, while near-infrared autofluorescence helped distinguish the graft region from healthy and atrophic tissue. Together, these observations support long-term structural persistence at the transplant site.
The study follows one person. It cannot estimate comparative efficacy, uncommon adverse events, or the likelihood that another patient will have the same course. The complete article was unavailable for review, so this capsule stays within its abstract and does not infer visual-acuity changes or a comprehensive safety history. The defensible milestone is sustained anatomical evidence ten years after transplantation, not proof that an iPSC graft restores sight.
2. Parkinson's cell manufacturing follows the product from fibroblast to neuronal precursor
Primary source: Cell Stem Cell article record and abstract
An autologous program needs to qualify a new biological starting material for every patient. A September 17 paper addresses that problem by comparing whole-genome sequences across donor fibroblasts, reprogrammed iPSCs, and their dopaminergic neuronal precursor products. It also introduces NeuriTest, an RNA-sequencing analysis intended to predict product quality from empirical animal data.
The abstract reports reproducible qualification across multiple donors, efficacy in a rodent Parkinson's model, and a nine-month GLP toxicology study. These are manufacturing and preclinical findings. They do not establish benefit in people, validate every future donor's product, or prove that immune matching eliminates all graft-related risks.
The useful development is a more explicit link between the cell's history, molecular identity, and preclinical behavior. Whether the transcriptomic test predicts outcomes for genuinely new donors remains a critical question. Complete methods and performance data were unavailable through a legitimate public route, so no sensitivity, failure rate, or clinical release threshold is inferred here. Disclosed interests include Aspen Neuroscience employment, equity, and advisory relationships.
3. Self-amplifying RNA extends the useful life of an iPSC experiment
Full analysis: Self-Amplifying RNA Keeps an iPSC Experiment in View
Primary sources: September 16 journal record and complete earlier manuscript
Self-amplifying RNA carries machinery that sustains expression in the cytoplasm. In this study, it delivered either a neuronal programming factor or optical reporters without requiring construction of a permanently modified iPSC line. A single Ngn2-RNA delivery, combined with a 48-hour selection step, produced cultures with more than 90% TUJ1-positive cells at day six.
For cardiac spheroids, researchers introduced the reporter before three-dimensional differentiation. Reporter signal remained usable beyond 30 days, allowing calcium activity to be observed as the tissue developed and after drug exposure. Selection and culture conditions were part of the workflow; a single transfection did not mean a single-step experiment.
The new event is publication in Cell Reports Methods. Our technical analysis uses the complete October 2025 manuscript, with that version boundary stated explicitly. The work supports a research tool for programming and observation. Early neuronal markers do not certify a transplant-ready product, and avoiding an integrating vector does not by itself establish genetic or clinical safety.
4. MSC vesicles protect retinal measurements, but visual behavior remains unresolved
Full analysis: MSC Vesicles Protect an Injured Mouse Retina, With a Mechanism Still to Test
Primary source: MedComm full article
A September 16 study tested human umbilical-cord MSC-derived extracellular vesicles in mice with sodium-iodate-induced retinal injury. Intravitreal administration and topical drops improved retinal electrical responses and structural measurements during a short experiment. The behavioral optomotor test, however, did not show a statistically significant improvement.
The authors connect protection to reduced lipocalin-2, a protein implicated in iron-linked oxidative injury, and nominate miR-486-3p as an active vesicle cargo. A reporter assay supported direct binding to the Lcn2 transcript. Delivering a synthetic miRNA agomir reproduced several protective measurements, providing evidence that this RNA can contribute to the effect.
The missing test is necessity: does removing miR-486-3p from the vesicles remove their benefit? The authors acknowledge that gap. Several central functional comparisons used only three animals per group, and an acute toxic injury does not reproduce slowly evolving human dry AMD. This is a mechanistic lead with a small-animal functional signal, not evidence that vesicle eye drops restore human vision.
5. MESEMS immune findings do not overturn the parent trial's MRI result
Primary source: Transplantation and Cellular Therapy article record and abstract
An ancillary analysis of the randomized, double-blind MESEMS trial reports peripheral immune measurements from 51 people with multiple sclerosis. The parent trial did not demonstrate a significant effect of autologous MSCs on MRI disease activity. This September 15 paper asks a different question: did the treatment measurably alter circulating immune-cell populations?
The abstract describes short-term shifts in B cells, regulatory T cells, and Th1/17 cells, alongside more limited longitudinal effects. Analyses used two distinct sample settings: freshly processed blood in a pilot cohort and cryopreserved peripheral blood mononuclear cells in the fuller cohort. Those measurements should not be treated as interchangeable replications of one result.
The findings support further investigation of biological activity and candidate biomarkers. They do not demonstrate remyelination, disability improvement, or a rescued efficacy result for MESEMS. Complete methods were unavailable during review; multiplicity handling, subgroup robustness, and the predictive performance of the exploratory associations therefore remain outside this capsule's claims. A treatment can alter blood immunology without producing the clinical effect it was developed to deliver.
6. A myelin model brings patient immune cells into a scalable assay
Primary source: September 18 bioRxiv preprint
Researchers describe a human myelin microphysiological system built by culturing neural organoids on directional microfibers in 3D-printed devices, then adding oligodendrocyte progenitors. The format supports 96 models in a conventional well plate. That is an assay-capacity figure, not a count of independent patients.
According to the abstract, T cells and monocytes from people with MS induced more demyelination than cells from healthy donors. Imaging and immune-cell measurements also separated healthy-donor, untreated-MS, responder, and nonresponder profiles after exposure to several MS-related drugs.
The platform could make immune-mediated myelin injury easier to compare across samples. Its title's promise of individualized treatment-response testing needs prospective validation: a model must predict outcomes in patients it has not already characterized. The full manuscript was not accessible during review, so donor counts, classification accuracy, validation design, and the meaning of each response label are not inferred here. This non-peer-reviewed preprint describes a disease-modeling approach, not an established test for choosing a patient's medication.
7. Laromestrocel preprint links inflammatory measurements with Alzheimer’s outcomes
Primary source: September 17 medRxiv preprint
A new laromestrocel manuscript reports inflammatory, brain-volume, cognitive, and blood-biomarker analyses in a study population of 49 people with mild Alzheimer’s disease. The abstract describes placebo comparisons across dose groups and associations between reduced inflammatory measurements, less atrophy, and cognitive scores. Individual comparisons involve substantially fewer participants than the total population.
These observations may help develop a mechanistic account of the MSC product. Correlation between two changing measures does not establish that suppressing inflammation caused a cognitive benefit, however. Multiple brain regions, time points, dose groups, and biomarkers also make prespecification and multiplicity handling important to interpretation.
The corresponding author lists Longeveron as an affiliation. The complete preprint was unavailable through the public routes checked, and the work has not been certified by peer review. This brief therefore reports the hypothesis and abstract-level observations without endorsing the title's causal claim, inferring the inflammation assay, or declaring a disease-modifying effect. A fuller assessment requires the analysis plan, missing-data accounting, complete safety results, and independent clinical confirmation.