A woman born with almost no functional vision because of a genetic mutation has regained measurable sight after a single CRISPR-based injection into her retina. The result comes from the BRILLIANCE phase 1/2 trial, an open-label, single ascending dose study that enrolled 14 participants with Leber congenital amaurosis type 10, a condition caused by a specific splice-site mutation in the CEP290 gene. The trial delivered a therapy called EDIT-101 through subretinal injection in one eye, and the early clinical data, published in the New England Journal of Medicine, represent the first time an in vivo CRISPR gene-editing treatment has produced functional vision gains in humans.
Why editing CEP290 inside the eye changes the treatment equation
Leber congenital amaurosis type 10 has resisted conventional gene therapy for a straightforward biological reason: the CEP290 coding sequence is too large to fit inside an adeno-associated virus vector, the standard delivery vehicle for retinal gene replacement. Preclinical work in animal models established this size constraint and proposed an alternative: rather than replacing the entire gene, use CRISPR to cut out the intronic mutation, known as CEP290 c.2991+1655A>G or IVS26, that causes aberrant splicing. By removing the mutation at the DNA level, the therapy aims to let retinal cells produce normal CEP290 protein on their own.
That strategy shifts the question from “can we deliver a working copy of the gene?” to “can we make precise enough edits inside living photoreceptor cells to restore their function?” The BRILLIANCE trial was designed to answer the second question. Its trial listing confirms the open-label, single ascending dose design targeting participants who carry the IVS26 genotype. Each participant received a single EDIT-101 subretinal injection in one eye, and the study tracked both safety and functional vision outcomes across escalating dose cohorts, with careful monitoring for retinal inflammation, off-target effects, and systemic exposure.
A key hypothesis emerging from these results is that the degree of vision restoration will depend not simply on how much vector reaches the retina but on how efficiently the CRISPR machinery edits the CEP290 mutation in individual retinal cells. If that relationship holds, future imaging or molecular assays that can measure editing frequency in vivo would become the most important predictor of patient outcomes, more informative than dose level alone. The trial data so far cannot fully resolve this question because patient-level editing efficiency has not been reported, but the functional improvements observed across dose cohorts offer early directional evidence that even partial correction in a subset of photoreceptors can translate into noticeable gains.
Fourteen participants, two pediatric cases, and measurable gains
The BRILLIANCE trial enrolled 14 participants with severe vision loss from Leber congenital amaurosis type 10, including pediatric patients treated at Children’s Hospital of Philadelphia. CHOP reported that two of its patients showed measurable improvements in vision after receiving the single subretinal injection, with better light perception and improved navigation in low-light environments. The NEJM paper documenting these results carries DOI 10.1056/NEJMoa2309915, and a news analysis in Nature noted that the improvements, while modest in absolute terms, exceeded what had been observed in earlier retinal gene therapy approaches for this specific condition.
The distinction between “modest” and “meaningful” matters. For someone who has lived with near-total blindness since birth, the ability to detect light, perceive shapes, or move through a room without assistance represents a qualitative shift in daily life. The clinical endpoints in the trial therefore emphasized measures such as full-field light sensitivity, navigational mobility tasks, and patient-reported visual function. Traditional letter-chart acuity, a mainstay of ophthalmology trials, is less informative here because many participants began with vision too poor to read any letters at all.
The inclusion of pediatric patients also raises a separate set of considerations. Younger retinas may retain more viable photoreceptor cells, which could make them more responsive to editing and potentially allow greater functional recovery. However, treating children with an irreversible gene-editing therapy demands a higher threshold of safety evidence and ethical scrutiny. The ascending dose design was structured in part to address this tension, starting with lower doses in adults to establish an initial safety profile before moving to higher doses and younger participants.
According to the Nature news coverage, several adult participants experienced improvements in light sensitivity and navigational ability, though not all responded to the same degree. This variability underscores the importance of baseline retinal health: patients with more surviving photoreceptors at the time of treatment are more likely to benefit, suggesting that earlier intervention could be advantageous but also more ethically complex.
Editing efficiency, durability, and what the data cannot yet show
Several questions remain open. The NEJM paper contains patient-level visual acuity and mobility data that have not been released through the ClinicalTrials.gov registry, limiting independent analysis of how outcomes varied across dose levels and age groups. The specific editing efficiency achieved in retinal cells-the percentage of target cells where the IVS26 mutation was successfully removed-has not been publicly reported for human participants. Without that number, it is difficult to establish whether the functional gains observed are close to a ceiling or whether higher editing rates, or alternative delivery vectors, could produce substantially better outcomes.
Durability is another gap. Preclinical studies in animal models, including non-human primates, suggested that CRISPR edits in photoreceptors should be stable over time because these cells are largely non-dividing. However, the long-term stability of edits in human retinas remains unconfirmed beyond the trial’s reported follow-up window. It will take years of continued observation to determine whether the restored function persists, plateaus, or gradually declines as disease processes continue to damage the retina.
Safety signals also require extended monitoring. To date, the trial has not reported serious adverse events clearly attributable to the editing itself, but late-emerging effects such as delayed inflammation, immune responses to Cas9, or unanticipated off-target edits cannot be ruled out. In the confined space of the eye, even localized inflammation can threaten remaining vision, so the risk-benefit calculation will depend heavily on how stable the safety profile looks as follow-up lengthens.
Regulators and clinicians will be watching not only for ocular side effects but also for any evidence of systemic exposure. Subretinal injection is designed to limit the spread of viral vectors beyond the eye, yet trace amounts can enter circulation. The absence of systemic complications so far is encouraging, but regulators are likely to require long-term surveillance, especially if CRISPR-based ocular therapies move into larger phase 3 studies or broader clinical use.
Implications for CRISPR medicine far beyond the retina
The BRILLIANCE results carry significance beyond this rare form of inherited blindness. They represent the first clear demonstration that in vivo CRISPR editing can produce functional benefit in humans, not just molecular corrections measured in a lab assay. For the broader field of genome editing, this offers proof of concept that carefully targeted edits in accessible tissues can translate into clinically meaningful outcomes.
At the same time, the trial highlights the limitations of current technology. Delivering CRISPR components efficiently and safely to enough cells remains a central challenge, especially in organs that are less accessible than the eye. The retina offers a relatively contained, immune-privileged environment and supports precise local delivery, conditions that are difficult to replicate in the liver, heart, or brain.
Another implication is the likely need for personalized or genotype-specific therapies. EDIT-101 is tailored to the IVS26 mutation in CEP290; it will not help patients with other forms of Leber congenital amaurosis or different CEP290 variants. Developing bespoke CRISPR treatments for many rare mutations raises questions about cost, regulatory pathways, and equitable access. Nonetheless, the success of a mutation-specific approach strengthens the case for creating modular platforms that can be rapidly adapted to new targets.
Nature’s reporting on the trial, including coverage accessible through the journal’s login portal, emphasizes this tension between scientific promise and practical constraints. Researchers see the BRILLIANCE data as a landmark but also as an early, imperfect prototype. Future generations of ocular gene editing may use smaller or transient nucleases, base editors, or prime editing systems to reduce off-target risks and expand the range of treatable mutations.
For now, the woman who can newly perceive light and shapes after a lifetime of darkness embodies both the hope and the uncertainty of CRISPR medicine. Her improved vision is a tangible, deeply personal outcome, yet it is also a data point in a cautious, stepwise effort to rewrite the genetic instructions inside living human cells. As longer-term results from BRILLIANCE emerge, they will help determine whether editing CEP290 in the eye becomes a durable therapy and, more broadly, whether in vivo CRISPR interventions can move from experimental milestones to reliable treatments.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
