This Week in Ophthalmology — Aug 22, 2026

Welcome to This Week in Ophthalmology. This week we're covering 10 notable papers spanning intravitreal drug choice and safety, the systemic disease that hides behind ocular findings, and new data on myopia control, geographic atrophy phenotyping, and surgical technique. Let's dive in.

We start with medical retina, where two papers ask different questions about the same crowded injection cabinet. In the British Journal of Ophthalmology, Liu and colleagues report a systematic review and frequentist network meta-analysis comparing high-dose aflibercept, 8 milligrams, against faricimab for macular oedema from retinal vein occlusion, using four-weekly aflibercept 2 milligrams as the common comparator across three trials and just over two thousand adults. There were no head-to-head trials, so everything here is indirect. Mean visual acuity gain did not differ between the agents. What did differ were the details clinicians actually negotiate with patients. Eight-weekly high-dose aflibercept, after five monthly loading doses, was about eighteen percent more likely to get an eye to the twenty-forty driving threshold than four-weekly faricimab. Faricimab, in turn, achieved complete resolution of intraretinal and subretinal fluid about twelve percent more often. And on safety, both aflibercept regimens carried a higher risk of a clinically significant pressure spike in the half hour after injection, with the high-dose regimens showing something on the order of a six to seven-fold greater risk than faricimab, though the estimates are imprecise. The practical read is that these agents are broadly equivalent for vision, and your choice should turn on whether you are chasing a dry macula, chasing a driving standard, or protecting a compromised optic nerve from post-injection pressure.

That safety thread continues in Retina, where Chhablani and colleagues interrogated the United States Food and Drug Administration Adverse Event Reporting System for retinal vasculitis events linked to intravitreal anti-VEGF agents and complement inhibitors. Between 2020 and 2025 they identified 554 reports attributable to a single agent, with reporting clearly rising after 2019. Brolucizumab dominated at just over half of all reports, followed by faricimab at about one in six, pegcetacoplan at roughly one in eight, and much smaller shares for the aflibercept and ranibizumab products. Only one report involved an aflibercept biosimilar and none involved a ranibizumab biosimilar. The authors are appropriately blunt about the ceiling on these data: spontaneous reporting cannot establish incidence, causality, or comparative risk, and reporting itself is shaped by publicity and litigation. Read alongside the network meta-analysis, the message is that comparative safety in this space remains genuinely unresolved, and that consenting patients for complement inhibition or newer bispecific therapy should include intraocular inflammation as a real, if unquantified, possibility.

The second theme is the eye as a window onto systemic disease and systemic drugs. In Retina, Hayat and colleagues followed 190 consecutive patients with central, branch, or cilioretinal artery occlusion for a median of just over two years. Major adverse cardiovascular and cerebrovascular events occurred in about fourteen percent of patients, and critically, roughly a quarter of those events happened within the first month after the occlusion. Hyperlipidaemia independently tripled the risk in both logistic and Cox models. Central occlusions carried higher risk than branch or cilioretinal events, though that reached significance in only one model. Visual outcome was driven by baseline acuity and age; eyes treated with hyperbaric oxygen improved within group, but final acuity did not differ significantly from untreated eyes, so this is not evidence of benefit. The actionable point is unchanged and worth repeating to every trainee: retinal artery occlusion is a stroke equivalent requiring same-day systemic workup, and lipid management is the modifiable lever.

On the pharmacological side, also in the British Journal of Ophthalmology, Oliver and colleagues used the TriNetX international database to examine fenofibrate in patients with diabetes and mild or moderate non-proliferative retinopathy, propensity-matching roughly four thousand fenofibrate users against non-users for each outcome. Fenofibrate use was associated with roughly half the odds of progressing to vision-threatening retinopathy, around forty percent lower odds of developing cataract, and about a two-thirds reduction in the odds of developing diabetic macular oedema. Primary open-angle glaucoma showed no significant association. This is retrospective claims-based work with all the confounding-by-indication that implies, but it aligns directionally with the randomized fenofibrate retinopathy data and strengthens the case for asking your endocrinology and primary care colleagues about fenofibrate in patients with early retinopathy.

Still in the British Journal of Ophthalmology, Chen and colleagues analysed nearly half a million UK Biobank participants and found that glaucoma was associated with roughly three times the odds of cataract after adjusting for demographics, comorbidities, and ocular parameters. Topical treatment carried similar odds, but participants who had undergone glaucoma surgery had about twelve times the odds of cataract. This is cross-sectional, so causal direction and surveillance bias are genuine problems, and the laser subgroup was too small to estimate reliably. Still, it quantifies what glaucoma surgeons already counsel: filtering surgery accelerates lens opacity, and these patients need earlier and more structured cataract surveillance.

Our third theme is phenotyping and prognostication. In the American Journal of Ophthalmology, de Carlo Forest and colleagues prospectively followed 81 eyes from 48 patients in a Colorado age-related macular degeneration registry, each with at least five fundus autofluorescence time points, using artificial intelligence segmentation with specialist review to model geographic atrophy growth. They describe three candidate phenotypes: unifocal foveal-involving, large coalescing multifocal, and small numerous multifocal. Both multifocal phenotypes had more severe growth trajectories and larger projected maximum lesion sizes than the unifocal foveal-involving pattern, and eyes with a peripapillary component grew faster and were projected to reach larger final sizes. The sample is small and single-centre, but in an era where complement inhibitors are judged on lesion growth rate, baseline phenotype could be a meaningful source of imbalance in trials and a useful prognostic signal in clinic. That sits naturally alongside a narrative review in Retina from Iglicki and colleagues, which synthesises the optical coherence tomography and angiography biomarker literature — disorganisation of the retinal inner layers, hyperreflective foci, fluid compartment patterns, perfusion change — into a framework for distinguishing angiogenic from inflammatory disease and for timing surgery in tractional maculopathy. It is a review, not new evidence, but it is a useful organising structure for how many of us already read a scan.

Ophthalmology brings a phenotyping paper from the neuro-ophthalmic side. Battista and colleagues describe 77 patients with late-onset Leber hereditary optic neuropathy, meaning onset at or after age forty, drawn from a larger Italian cohort of 398. The m.11778G-to-A variant predominated, the male predominance was less pronounced than in classic disease, close to sixty percent of patients had a smoking history, and more than four in five affected women were postmenopausal. Overall visual recovery was just under forty percent, and an exploratory decision-tree analysis linked recovery to idebenone at 900 milligrams a day or higher and to J or T mitochondrial haplogroups. These are hypothesis-generating associations from a retrospective cohort, not treatment evidence, but they support asking about smoking and hormonal status in later-life optic neuropathy and not dismissing the diagnosis on age alone.

Finally, two papers about what we put in front of, and inside, the eye. In JAMA Ophthalmology, Shen and colleagues report the 24-month results of a double-masked randomized trial of spectacle lenses with highly aspherical lenslets versus single-vision lenses in 159 United States children aged six to twelve, across nine sites, with about eighty-five percent completing follow-up. Myopia progression was reduced by roughly seventy percent in relative terms, a difference of about two thirds of a dioptre over two years, and axial elongation was cut by roughly half, a difference of about a quarter of a millimetre. Best-corrected acuity was equivalent, ocular adverse events were similar between groups and none were lens related, and symptom reports were if anything fewer in the treated group. This is the trial underpinning Food and Drug Administration authorisation of the device, and it extends efficacy previously shown mainly in Asian cohorts to a United States population. And in Retina, Cho and colleagues compared transscleral four-point fixation using a closed continuous-loop suture against the Yamane technique across 179 eyes. Visual and refractive outcomes at six months were comparable, and complication rates including corneal decompensation, hypotony, and endophthalmitis did not differ. Yamane was about fourteen minutes faster, but carried a substantially higher hazard of lens redislocation, with a point estimate around eight-fold and a very wide range around it given few events. Speed versus durability, and the authors argue preoperative structural assessment should drive that choice.

If you only have time for one paper this week, make it the JAMA Ophthalmology trial of highly aspherical lenslet spectacles in United States children. It is a randomized, double-masked, regulator-grade result in a population where these lenses were previously unproven, and it changes what you can offer a myopic child at the next visit.

Here are the key takeaways from this week in Ophthalmology. First, highly aspherical lenslet spectacles substantially slowed myopia progression and axial elongation in United States children over two years with no safety signal. Second, high-dose aflibercept and faricimab give similar visual gains in vein occlusion oedema; pick on drying, driving vision, or post-injection pressure risk, and remember all these agents carry retinal vasculitis reports in pharmacovigilance data. Third, treat every retinal artery occlusion as an acute systemic emergency, with the highest event risk in the first month and hyperlipidaemia as the strongest modifiable predictor. Fourth, fenofibrate in early diabetic retinopathy is associated with less progression, less macular oedema, and less cataract, worth a conversation with the prescribing physician. And fifth, both glaucoma and especially glaucoma surgery are strongly associated with cataract, and geographic atrophy phenotype on autofluorescence predicts how fast the lesion will grow.

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.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And a quick rating on Apple Podcasts or Spotify helps other physicians discover the show.

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