This week in Ophthalmology
Nov 22, 2026
Grafting vs Intraoperative Mitomycin C for Preventing Pterygium Recurrence: A Randomized Clinical Trial.
In 750 adults randomized after primary pterygium excision, one-year recurrence was about 5 percent with conjunctival autograft, 14 percent with mitomycin C and 23 percent with amniotic membrane.
JAMA Ophthalmology · 2026 · PubMed
This week’s highlights
Read the weekly briefing
This AudioScholar briefing is generated by artificial intelligence from PubMed abstracts for healthcare professionals and is not medical advice.
This week: conjunctival autograft outperforms mitomycin and amniotic membrane against pterygium recurrence, a stronger atropine dose for myopic children with intermittent exotropia, and how complement inhibitors for geographic atrophy are used once they leave the trials.
Welcome to This Week in Ophthalmology. We're covering 10 papers this week: 3 in depth, 4 in brief, 3 quick signals.
Writing in JAMA Ophthalmology, Peng and colleagues report a multicentre, open-label randomized trial with masked outcome assessment, run at five tertiary hospitals in China [1]. It enrolled 750 adults with primary pterygium, all of whom had a standardized excision and were then randomized equally to an amniotic membrane transplant, intraoperative mitomycin C, or a conjunctival autograft. The primary outcome was recurrence at twelve months. Recurrence was about 5 percent with conjunctival autograft, about 14 percent with mitomycin C, and about 23 percent with amniotic membrane, and every pairwise difference was statistically significant. Put another way, recurrence after amniotic membrane was more than four times the rate seen after autograft. Adverse events were uncommon in all three arms, with no intervention-related serious events, and over nine in ten participants completed follow-up. The main limitations are the open-label surgery, a single country, and only one year of follow-up, which may miss later recurrences. The authors describe the differences as moderate and frame recurrence as an important but not exclusive factor in choosing a technique. What it adds is a large, three-way, randomized ranking of the common adjuncts, where most prior comparisons were smaller or pairwise.
Also in JAMA Ophthalmology, Wang and colleagues report the AMIXT2 trial, a single-centre, double-masked, three-arm randomized trial in China [2]. It enrolled 264 children aged six to twelve with myopia and basic-type intermittent exotropia, randomized to nightly 0.01, 0.02 or 0.04 percent atropine in both eyes. Over twelve months, myopia progressed by about 0.43 dioptres on the lowest dose and about 0.18 dioptres on 0.04 percent, so the highest concentration slowed progression by roughly a quarter of a dioptre. Axial elongation on 0.04 percent was about half that seen on 0.01 percent. Near exodeviation also fell further on the highest dose, by about six prism dioptres against about three and a half, a finding the authors flag as needing further study. Photophobia affected about one in five children on 0.04 percent, roughly double the lower doses, though no child stopped treatment because of it. Limitations include a single centre, about 15 percent of children missing the twelve-month visit, one year of data and no placebo arm. The trial adds dose-response evidence in a subgroup where the effect of atropine on ocular alignment has been a concern.
In Ophthalmology Retina, Maturi and colleagues describe how pegcetacoplan and avacincaptad pegol for geographic atrophy are used in routine practice [3]. This retrospective, multicentre cohort drew on the Vestrum Health Retina Database and included over twenty thousand eyes from about fourteen and a half thousand patients. Early retreatment was frequently delayed: only 58 percent of pegcetacoplan-only eyes had a second injection within sixty days, and fewer than one in five avacincaptad-only eyes had a second injection within thirty-five days. About a fifth of eyes had a ninety-day gap in care during the first year. Among eyes without baseline neovascular disease, about 6.6 percent converted to neovascular AMD within twelve months, and avacincaptad monotherapy was associated with modestly lower odds of conversion than pegcetacoplan. Eyes with at least six injections lost about four letters on average over the year. Endophthalmitis occurred after fewer than two in every thousand injections, and intraocular inflammation was rare, but more than half of those inflammation cases lost thirty letters or more. As an electronic health record study with no untreated comparator, it cannot separate drug effects from patient selection.
Now, four in brief.
A network meta-analysis in Survey of Ophthalmology from Butt and colleagues pooled 13 randomized trials with 5,126 patients to compare the same two complement inhibitors indirectly [4]. With moderate certainty, pegcetacoplan given monthly or every other month slowed lesion growth against sham, and avacincaptad doses did so when the trials' own model-based estimates were used. No regimen of either drug differed from the other on lesion growth, and no complement inhibitor improved low-luminance visual acuity. Both pegcetacoplan regimens and avacincaptad 4 milligrams increased neovascular AMD against sham, with lower incidence on avacincaptad 2 milligrams than monthly pegcetacoplan, which points the same way as the registry signal above. That comparison is indirect and low certainty.
Staying with intravitreal safety, Yasuda and colleagues report in Ophthalmology Retina on faricimab-induced intraocular inflammation from the J-CREST group [5]. The inflammation was rare, affecting just under half a percent of patients and roughly one in a thousand injections. It presented either as anterior-limited inflammation or as intermediate uveitis. Intermediate uveitis caused greater visual decline, yet vision recovered with appropriate corticosteroid treatment. These are observational data describing a rare event, so the estimates of incidence and recovery rest on a small number of affected eyes.
In Ophthalmology, Huang and colleagues studied 5,687 children at a tertiary children's hospital who had intraocular pressure measured within sixty days of starting any steroid [6]. Nearly one in five children developed ocular hypertension. Pre-existing glaucoma carried close to seven times the odds, prednisone more than tripled the odds relative to dexamethasone, and ocular and intravenous routes carried roughly three and a half and two and a half times the odds of topical use. Uveitis and juvenile arthritis added modest risk, while age, sex, diabetes and asthma did not. This is a single-centre retrospective cohort of children who were selected for pressure checks.
From Cornea, Aguwa and colleagues analysed nearly three hundred thousand patients with infectious keratitis in the IRIS Registry [7]. Treatment failure, defined as keratoplasty or removal of the eye within ninety days, occurred in about 1.3 percent. Presenting vision was the strongest predictor: vision worse than 20/400 carried about twenty-six times the odds of failure compared with 20/40 or better. Older age, male sex, Black race, Hispanic ethnicity and Medicare insurance carried smaller increases, and hypopyon and perforation carried high failure rates. Registry coding and a surgery-based definition of failure limit precision, but the scale is unusual.
Three quick signals.
In Ophthalmology, Shah and colleagues matched 3,695 emergency department pairs and found that patients with retinal artery occlusion had brain imaging about 77 percent of the time, against about 92 percent for transient ischemic attack, with slightly more cervical imaging [8]. The abstract reports imaging use only, not patient outcomes.
A narrative review in Survey of Ophthalmology from Palia and colleagues covers rhegmatogenous retinal detachment, noting that over 90 percent of eyes reattach, yet up to 40 percent of reattached eyes do not recover functional vision [9]. Proliferative vitreoretinopathy and access to care feature prominently. It is a synthesis, not new data.
In Progress in Retinal and Eye Research, Smith and colleagues review retinopathy of prematurity and its shift toward prevention [10]. They report that early enteral omega-3 and omega-6 supplementation reduces neovascularization risk, that timed topical corticosteroids show promise, and that risk algorithms with near-complete sensitivity can reduce screening examinations. It is a narrative review.
Stepping back across this week's papers, much of the evidence comes from large registries and health record databases, which describe practice and associations at scale but cannot separate a treatment's effect from the patients who received it. The two randomized trials were each confined to one country and one year. For complement inhibitors, trial synthesis and real-world data point the same way on neovascular conversion, yet both depend on indirect comparison, and direct head-to-head evidence is still absent.
If you only have time for one paper this week, make it the pterygium trial from Peng and colleagues [1]. It gives a large randomized ranking of three widely used adjuncts that had largely been compared in smaller or pairwise studies.
That's your roundup for This Week in Ophthalmology. The full transcript and references are available on the episode page. See you next week.
If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And think of one colleague, in any specialty, who never has time to keep up with the literature. Tell them about AudioScholar: a ten-minute weekly for every specialty, to listen to in any podcast app, or to read at audioscholar dot C C.
This is an automated summary generated by artificial intelligence, which can make mistakes. Use it to choose which papers to read in full.
Get Ophthalmology every week
By email, Spotify, Apple or any podcast app.
Spot something worth flagging?