This Week in Ophthalmology — Jun 12, 2026

Welcome to This Week in Ophthalmology. This week we're covering 9 notable papers spanning new therapeutic approaches in retina, refinements in managing uveitis, and how new technology is changing risk stratification and diagnosis. Let's dive in.

We begin with two studies offering hope for retinal diseases with limited treatment options. First, in The New England Journal of Medicine, we have early results from a gene therapy trial for X-linked retinoschisis [1]. This study involved a single subretinal injection of an AAV8 vector into one eye of 12 patients aged 5 to 18. Over 52 weeks, the therapy demonstrated a good safety profile, with no grade 3 or higher adverse events or ocular inflammation reported. Structurally, the results were striking: swept-source OCT showed closure of the macular schisis cavity in all 12 treated eyes by week 13, with a mean reduction in central retinal thickness of over 400 microns. Functionally, the mean best-corrected visual acuity improved by 10.8 letters in the treated eyes, compared to just 2.4 letters in the untreated fellow eyes. However, the investigators observed no clinically meaningful changes in photoreceptor and bipolar cell function on full-field ERG or in macular sensitivity on microperimetry. While this is a small, early-phase study, the anatomical and visual acuity improvements are encouraging and warrant further clinical testing. Also this week, The British Journal of Ophthalmology published a multicentre randomised trial on photobiomodulation, or PBM, for early and intermediate dry age-related macular degeneration [4]. In this double-masked study, 138 eyes were randomised to either PBM or a sham treatment. The primary outcome was the change in mean drusen volume over 12 months. The PBM group saw a significant decrease in drusen volume, while the sham group experienced an increase. This translated to a functional benefit as well: visual acuity improved by an average of 1.3 letters in the PBM group, but declined by 2.6 letters in the sham group, resulting in a statistically significant between-group difference of nearly 4 letters. On the safety front, no retinal phototoxicity was observed. Interestingly, four eyes in the sham group developed macular neovascularisation, compared with none in the PBM group, a difference that reached statistical significance. This study suggests that PBM may not only reduce drusen burden and improve vision, but could also potentially slow progression, making it a promising therapeutic option for a large patient population with few alternatives.

Turning to inflammatory eye disease, three papers this week provide new insights into diagnosis and management. First, a clinically important study from the American Journal of Ophthalmology investigated the impact of anti-adalimumab antibodies in patients with uveitis [5]. In a retrospective cohort of 128 patients, researchers found that the development of these antibodies is common, with a cumulative incidence approaching 50% by 8 years of therapy. The presence of antibodies had a clear biological effect, being associated with significantly lower serum adalimumab concentrations. Most critically for clinical practice, anti-adalimumab antibodies were strongly associated with active ocular inflammation, increasing the odds by more than threefold. The study also hinted that weekly dosing, as opposed to every other week, and concomitant immunosuppression might reduce the risk of antibody formation. This paper provides strong evidence to support testing for anti-drug antibodies in patients who experience breakthrough inflammation while on adalimumab. Next, from Ophthalmology. Retina, a large retrospective study helps clarify the distinct clinical phenotypes of herpesvirus retinitis based on the host's immune status [8]. The study analyzed 120 patients with either acute retinal necrosis, or ARN, or cytomegalovirus retinitis, known as CMVR. The findings confirm that CMVR is predominantly a disease of the immunocompromised, occurring mainly in patients with lymphoma or those on immunosuppressive therapy, and it tends to follow a more indolent course. In contrast, ARN was more likely to be unilateral, involve peripheral necrosis, and have a more fulminant course. The risk of retinal detachment was dramatically different between the two: ARN was associated with an over 11-fold increased risk of retinal detachment, which also occurred earlier. This work underscores how a patient's immune status shapes the disease presentation and prognosis, which can guide early recognition and management strategies. Finally, in The British Journal of Ophthalmology, researchers propose a new set of ocular-centric diagnostic criteria for Behçet's uveitis, developed and validated in a Chinese population [9]. The new criteria, called BU-SDC, use a weighted scoring system that prioritizes ocular findings. Key signs like vitreous cells or haze and retinal vasculitis on fluorescein angiography each score two points, while other signs like hypopyon or retinal hemorrhages score one point. A patient with an ocular score of 5 or more could be diagnosed with Behçet's uveitis, provided other specific conditions are excluded. If the ocular score is lower but suspicion remains high, systemic features like oral ulcers are then incorporated. In a validation cohort, these new criteria significantly outperformed both the International Study Group and the Standardization of Uveitis Nomenclature criteria for diagnostic accuracy. This framework could provide a more robust and validated method for diagnosing Behçet's uveitis, especially in cases where ocular signs are the primary presentation.

Our final theme explores how genetics, advanced imaging, and big data are shaping our understanding of ophthalmic and systemic disease. A major study in Ophthalmology demonstrates the power of a polygenic risk score, or PRS, for glaucoma [2]. Analyzing data from over 400,000 individuals in the FinnGen biobank, researchers found that a PRS could effectively stratify individuals by their lifetime risk of developing glaucoma. The effect was substantial: the lifetime risk by age 85 ranged from just 2.5% for those in the lowest percentile of risk to over 45% for those in the top percentile. This genetic risk was independent of family history, providing a more precise risk assessment. Furthermore, a high PRS was associated with a more severe disease course, predicting a greater need for medication escalation, laser therapies, and incisional surgery over a 20-year follow-up. These findings strongly support the potential utility of using genetic risk scores for targeted glaucoma screening and personalized management. From genetics to advanced imaging, a study in JAMA Ophthalmology explores the use of OCT angiography as a potential noninvasive biomarker for Alzheimer's disease [3]. In a cross-sectional study of 103 individuals with either normal cognition, mild cognitive impairment, or Alzheimer's dementia, researchers identified distinct retinal microvascular signatures. They found that the ganglion cell complex was thinner in patients with Alzheimer's compared to controls. Retinal vessel skeleton density progressively decreased with worsening cognitive status. Interestingly, choriocapillaris flow deficits showed a biphasic pattern, suggesting an initial compensatory hyperperfusion in mild cognitive impairment, followed by perfusion failure in later-stage dementia. While based on a relatively small cohort, these findings suggest OCTA could one day serve as an accessible screening tool for cognitive neurodegeneration, warranting larger longitudinal studies. Shifting to big data and social determinants of health, a paper in Retina provides a stark look at diabetic retinopathy outcomes among unhoused individuals [6]. Using a large electronic health records network, the study compared over 9,600 unhoused patients with diabetes to over a million housed controls. At initial presentation, unhoused individuals had dramatically higher rates of severe disease, including a more than three-fold higher rate of proliferative diabetic retinopathy and a four-fold higher rate of vitreous hemorrhage. They were also far less likely to maintain at least six months of ophthalmic follow-up. After matching for other factors, being unhoused independently increased the one-year risk of developing proliferative disease, diabetic macular edema, and vitreous hemorrhage. This powerful analysis identifies unhoused status as a critical, independent risk factor for poor outcomes, underscoring the urgent need for targeted screening programs and socially informed ophthalmic care. Finally, looking to the future of clinical decision support, a study in The British Journal of Ophthalmology evaluated the reasoning capabilities of multimodal large language models in ophthalmology [7]. Researchers tested models like ChatGPT-5 on 316 complex questions that paired a clinical vignette with an image. They found that prompting the models to explain their reasoning step-by-step was associated with numerically higher accuracy scores across both English and Chinese datasets. While ChatGPT-5 performed best, the study also revealed significant limitations in the models' subspecialty knowledge and image interpretation abilities. The work highlights both the potential of these AI tools and the critical need for rigorous evaluation of their reasoning process before they can be safely applied in educational or clinical settings.

If you only have time for one paper this week, make it the study on anti-adalimumab antibodies in the American Journal of Ophthalmology [5]. It shows that nearly half of patients on long-term adalimumab develop antibodies, which are strongly linked to active inflammation. This provides a strong rationale for testing for these antibodies when a patient's uveitis flares.

Here are the key takeaways from this week in Ophthalmology. First, for patients on long-term adalimumab with breakthrough uveitis, consider checking for anti-drug antibodies, as they are common and strongly associated with treatment failure [5]. Second, a polygenic risk score can powerfully stratify lifetime glaucoma risk, independent of family history, and predicts the need for more intensive therapy, which may guide future screening protocols [2]. Third, in patients with early or intermediate dry AMD, photobiomodulation appears to reduce drusen volume and improve vision over 12 months, offering a potential new treatment option for a condition with few therapies [4]. Fourth, remember that unhoused status is an independent and potent risk factor for severe diabetic retinopathy and poor follow-up, highlighting a critical need for targeted outreach and care models to address this health disparity [6]. And finally, in viral retinitis, the host's immune status drives the clinical picture: expect a more indolent course with CMVR in immunocompromised patients, and a more fulminant course with a high risk for retinal detachment with ARN [8].

That's your roundup for This Week in Ophthalmology. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.