This Week in Ophthalmology — Sep 12, 2026

Welcome to This Week in Ophthalmology. This week we're covering 10 notable papers spanning targeted immunotherapy for thyroid eye disease, medical and surgical retina — from complement inhibition to prophylactic laser — and a cluster of studies on ocular inflammation, infection and drug-induced surface disease. Let's dive in.

We start with thyroid eye disease, where a new biologic mechanism has now been tested in two large phase 3 trials. In Ophthalmology, Briceño and colleagues report the week 24 results of SatraGO-1 and SatraGO-2, identically designed, double-masked, placebo-controlled trials of subcutaneous satralizumab, an anti-interleukin-6 receptor antibody, in just over two hundred and fifty participants with active or inactive thyroid eye disease [1]. Among participants with active disease, the primary endpoint of a two millimetre or greater proptosis reduction was reached by about half of those on satralizumab in both trials, compared with roughly a quarter on placebo in SatraGO-2 — a statistically significant difference there — but the same comparison in SatraGO-1 was against a much stronger placebo response of about a third and did not reach significance. That is an important caveat: one of the two pivotal trials missed its primary endpoint. The secondary outcomes were more consistent, with clinical activity score reduction, disease inactivation, and diplopia improvement all favouring satralizumab, and the diplopia benefit reaching significance in SatraGO-2 only. Tolerability was the standout feature — low rates of serious adverse events and discontinuations, and no serious infections reported. For clinicians who have been weighing teprotumumab's hearing and hyperglycaemia risks, an interleukin-6 pathway agent given subcutaneously is a meaningful addition to the conversation, though the mixed proptosis result means enthusiasm should be calibrated. That mixed picture is precisely the problem taken up in JAMA Ophthalmology by Kazim and colleagues, who argue in an Innovate TED Initiative review that our current metrics — proptosis in millimetres, the clinical activity score, categorical diplopia grading — fail to capture the clinical complexity and diversity of thyroid eye disease, and may be limiting our ability to judge these new targeted agents at all [2]. They outline a programme of prospective validation studies to build better outcome measures. Read alongside the satralizumab data, it is a reminder that when a trial narrowly misses on proptosis, the failure may be partly one of measurement rather than of drug.

Moving to medical retina, two papers address how we might slow or prevent irreversible damage. In the American Journal of Ophthalmology, Shen and colleagues report a two-year real-world experience with intravitreal pegcetacoplan for geographic atrophy in seventy-seven eyes followed with swept-source optical coherence tomography angiography [3]. Eyes received an average of eighteen injections over two years, and the square-root geographic atrophy growth rate fell by about forty-one percent compared with each patient's own pre-treatment growth rate, with the benefit accumulating rather than waning in the second year — nearly every eye either held steady or slowed further. But the functional story is sobering: mean best-corrected visual acuity still declined by roughly eight letters over the two years, with no difference between foveal and non-foveal atrophy. Interestingly, eyes with faster pre-treatment growth showed greater percentage inhibition, which may help with patient selection. The practical message is to keep counselling realistic — you are buying anatomy, and vision continues to fall. Prevention gets a different angle in Ophthalmology, where Bison and colleagues used a large United States federated health records network to run propensity-matched cohorts of adults with type 2 diabetes taking cetirizine, ibuprofen, or prednisone [4]. Each of the three anti-inflammatory agents was associated with substantially lower three-year incidence of diabetic macular oedema and proliferative retinopathy — roughly a third to a half lower, depending on drug and outcome — with effect sizes comparable to the fenofibrate positive control, while gabapentin, the negative control, showed no association at all. The design is careful, with matching on forty-four covariates, negative-control outcomes, E-values and nine sensitivity analyses. Still, this is observational pharmacoepidemiology, and nobody should start prednisone to protect a retina. What it does justify is prospective evaluation of cheap anti-inflammatory agents as adjuncts in diabetic eye disease.

Turning to surgical retina, three papers cover prophylaxis, technique, and language. In Ophthalmology Retina, Khatami and colleagues pooled nine observational studies covering just over five thousand fellow eyes to ask whether prophylactic retinopexy prevents detachment in the fellow eye after a contralateral retinal detachment [5]. Detachment occurred in about four and a half percent of treated fellow eyes versus about seven and a half percent of observed eyes, and pooled analysis showed treatment cut the risk by more than half, with an even stronger point estimate when restricted to laser retinopexy and no heterogeneity between studies. The caveats are substantial — every included study was observational, the GRADE certainty was low overall and very low for the lesion-subtype analyses, and neither the lattice-degeneration-only nor the lattice-with-tear subgroups reached statistical significance on their own. So this supports discussing prophylaxis in selected high-risk fellow eyes, and supports someone finally running the randomised trial, but it does not settle which lesions to treat. On technique, in Retina, Seyyar and colleagues describe macular hole repair using a temporal inverted internal limiting membrane flap secured with dispersive viscoelastic under air, with air as the sole tamponade and no postoperative head positioning at all [6]. Across seventy-eight eyes, every single hole closed after one operation, mostly with a U-shaped configuration, and visual acuity improved significantly with no meaningful intraocular pressure change. This is a single-centre, uncontrolled series, so the one hundred percent closure rate needs replication, but for patients who cannot posture or who need to fly, an air-only, positioning-free approach is worth knowing about. And in Ophthalmology Retina, Couturier and colleagues present a Myopia Society consensus proposing a new nomenclature for myopic traction maculopathy [7]. Rather than a numbered classification, lesions are named in descending order of severity — macular detachment, then full-thickness macular hole, then lamellar hole, then retinoschisis — so a complex eye is described as, for example, myopic macular detachment with full-thickness hole and schisis. The panel also standardised optical coherence tomography protocols and tested multi-observer agreement. It will not change what you do in theatre tomorrow, but it should make referral letters and trial cohorts far more comparable.

Our last theme is ocular inflammation, infection and iatrogenic surface disease. In Retina, Farrell and colleagues reviewed two hundred and seventy-five eyes receiving the 0.18 milligram fluocinolone acetonide implant for non-infectious posterior segment uveitis [8]. Ocular hypertension above 24 millimetres of mercury occurred in about one in ten eyes over a mean follow-up near twenty months, and among those eyes a substantial proportion settled with observation alone while roughly one in ten needed incisional surgery. The strongest predictors were higher baseline pressure and already being on pressure-lowering therapy, which raised the odds several-fold. That is a reassuring safety profile overall, with a clear signal about who needs closer monitoring. In the American Journal of Ophthalmology, Veldhuis and colleagues report a prospective cohort from the Dutch BioDay registry, examining one hundred and fifty adults with moderate-to-severe atopic dermatitis starting dupilumab [9]. Strikingly, ocular surface disease was already present in ninety-four percent of patients at baseline, yet only about sixty percent of them reported any ocular symptoms — so most of this is silent until you look. Over twenty-eight weeks, about thirty-one percent of patients developed dupilumab-associated surface disease despite ophthalmic treatment use rising from roughly fifteen percent to about sixty-five percent, and impression cytology showed stable but low goblet cell numbers with significantly reduced mucin 5AC production. The practical implication is baseline ophthalmic examination before starting dupilumab, not symptom-triggered referral. Finally, in the British Journal of Ophthalmology, Burton and colleagues report a multi-institutional retrospective cohort of one hundred and eighteen eyes with endogenous fungal endophthalmitis across eight academic centres in the United States [10]. Nearly two thirds of patients started on intravenous rather than oral systemic antifungals, and that decision was driven by systemic features — symptoms and positive blood cultures — rather than by anything seen on ocular examination. Intravenous therapy was associated with less frequent surgery on unadjusted analysis, but that difference lost statistical significance after controlling for confounders. What did predict poor final vision was worse presenting acuity and infection with a mould rather than a yeast, which roughly quadrupled the odds of a bad outcome.

If you only have time for one paper this week, make it the SatraGO phase 3 report in Ophthalmology [1]. It introduces a subcutaneously delivered interleukin-6 receptor blocker into a therapeutic area that has been dominated by one expensive intravenous agent with a difficult safety profile — and the split result across the two trials is exactly the nuance you will need when patients ask about it.

Here are the key takeaways from this week in Ophthalmology. Satralizumab improved clinical activity score, disease inactivation and diplopia in thyroid eye disease with a clean safety record, but the proptosis primary endpoint was significant in only one of the two trials. Two years of pegcetacoplan slowed geographic atrophy growth by around forty percent, with the effect accumulating in year two, while visual acuity continued to decline — counsel accordingly. Common anti-inflammatory drugs were associated with markedly lower rates of vision-threatening diabetic retinopathy in a large matched cohort, a hypothesis for trials, not a prescribing change. Prophylactic retinopexy in fellow eyes after contralateral detachment more than halved subsequent detachment risk, on low-certainty observational evidence. And examine the eyes of every atopic dermatitis patient before starting dupilumab, because surface disease is nearly universal and largely asymptomatic at baseline.

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 if you have a paper you have been meaning to read, upload the PDF, or paste any link, at audioscholar dot C C. We will turn it into audio like this one, in any of thirty-one languages.

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