This Week in Ophthalmology — Aug 1, 2026

Welcome to This Week in Ophthalmology. This week we're covering 10 notable papers spanning vitreoretinal surgery and its downstream complications, medical and laser management of macular disease, and access and quality issues in corneal care. Let's dive in.

We'll start with a group of studies that all circle the same question: what happens to the retina after we intervene in the vitreous cavity. In JAMA Ophthalmology, a very large retrospective cohort using the TriNetX United States network asked whether a prior pars plana vitrectomy independently raises the risk of cystoid macular edema after cataract surgery [4]. Working from more than six hundred thousand cataract patients and excluding those with pre-existing macular edema or known risk factors, the investigators propensity-matched just over seven thousand pairs. Cystoid macular edema in the thirty to ninety day window occurred in about four and a half percent of previously vitrectomized eyes versus a little over one percent of controls — roughly a three-and-a-half to four-fold higher risk, and an absolute difference of about three percentage points. The signal held whether the original vitrectomy was for retinal detachment or for other indications, and it persisted after excluding eyes with intraoperative or postoperative cataract surgery complications. The authors are appropriately cautious: these are coding-based diagnoses with no visual acuity outcomes, so this does not tell you that prophylaxis works. But it does tell you that when you consent a vitrectomized eye for cataract surgery, a several-fold higher edema risk belongs in that conversation. Alongside that, Ophthalmology Retina published a related database analysis, again in TriNetX, looking at whether treating a retinal break without detachment sets the stage for epiretinal membrane [3]. Comparing roughly twenty thousand propensity-matched laser retinopexy patients and seven hundred and thirty cryotherapy patients against untreated controls with retinal breaks, membrane formation was about three times more common after laser at three months — just under three percent versus about one percent — and remained roughly two-and-a-half to three-fold elevated at one year, with higher membrane peel rates persisting out to five years. Cryotherapy showed a very similar pattern, and head-to-head there was no difference between the two modalities. The authors read this as membrane formation being a response to vitreoretinal disruption and inflammation rather than to any particular instrument. The obvious confounder is that treated breaks are probably different breaks — larger, more symptomatic, more likely to have vitreous traction — so causal language should be resisted. Still, taken together with the JAMA Ophthalmology paper, there's a coherent theme: eyes we have operated on or lasered carry a durable elevated risk of macular interface and macular fluid problems, and follow-up should reflect that.

Staying in the vitreous cavity, two papers deal with how we manage the perioperative period and the fellow eye. In Retina, a retrospective single-centre comparison of one hundred and seventy-two eyes tested a dropless regimen at the close of pars plana vitrectomy — intravitreal decanted triamcinolone five milligrams plus moxifloxacin five hundred micrograms — against standard topical moxifloxacin and prednisolone [2]. The dropless cohort had significantly less anterior chamber inflammation at two weeks and three months, lower central subfield thickness at two weeks, three months and six months, less cystoid macular edema, and better visual acuity gains — a fifteen letter improvement versus a two-and-a-half letter loss at two weeks, and twenty-five versus nine-and-a-half letters at three months. Intraocular pressure was similar and there were no cases of endophthalmitis. Those are large visual differences for a perioperative regimen change, and the design demands caution: this is a consecutive retrospective series with a historical control cohort from the preceding six months, so surgeon learning, case mix and secular trends are all in play, and eighty-seven eyes is far too few to say anything meaningful about endophthalmitis risk. It's a hypothesis worth a randomized trial rather than a settled answer. Also in Retina, a retrospective cohort from Sheba Medical Center followed seventy-seven patients with non-traumatic, non-syndromic giant retinal tears to see what happened to the fellow eye [8]. The numbers are sobering: retinal tears developed in forty-four percent of contralateral eyes, and about a third progressed to detachment or a giant tear of their own. Prophylactic three-hundred-and-sixty degree laser was performed in only twelve patients, and detachment rates were essentially identical in lasered and non-lasered eyes — thirty-three percent versus thirty-four percent. No association emerged with age, axial length or refraction. The authors' interpretation is not that prophylaxis is futile but that a two-line anterior barrier as they applied it is inadequate, and that denser, more posteriorly defined treatment would need prospective evaluation. The practical message for now is that the fellow eye in giant retinal tear disease deserves intensive, ongoing surveillance and a frank discussion that a light barrier laser may give false reassurance.

Our second theme is medical and laser therapy for macular disease, and here Ophthalmology Retina and Retina between them cover three quite different interventions. First, a target trial emulation from the Mass General Brigham database asked whether sustained high-intensity lipophilic statin use — atorvastatin or simvastatin at forty to eighty milligrams — is associated with lower rates of age-related macular degeneration [5]. Among hypertensive adults fifty-five and older, propensity-matched cohorts of roughly eight thousand six hundred per arm showed incident non-exudative macular degeneration in under one percent of exposed versus about one percent of unexposed over an average of under three years, a roughly forty-seven percent lower hazard. In a separate cohort of just over a thousand matched pairs who already had non-exudative disease, progression to exudative disease occurred in about six percent of exposed versus nearly eight percent of unexposed, roughly a forty percent risk reduction. Importantly, statin use of any kind showed no association, which the authors interpret as a regimen-specific effect. The negative control, posterior vitreous detachment, showed no association, which is reassuring; the positive control, cataract, moved in the expected direction. But event counts are small, follow-up is short for a disease that takes decades, and outcomes are clinically documented diagnoses. This is not a reason to start a statin for retinal indications — it is a reason to support the prospective trials the authors call for. Two papers tackle central serous chorioretinopathy from opposite ends. In Retina, a systematic review of six randomized trials totalling four hundred and one participants with acute disease found that subthreshold laser probably produces similar visual acuity to conventional laser with less retinal pigment epithelium damage, but — and this is the part to hold onto — probably a lower rate of subretinal fluid resolution [7]. Against observation, and against half-dose photodynamic therapy, the evidence was of very low certainty, with subthreshold laser showing comparable vision but lower anatomical improvement than photodynamic therapy. So subthreshold laser is safer, but its anatomical efficacy in acute disease is limited, and its advantage over simply watching remains unproven. Meanwhile, for chronic disease, a pilot open-label randomized trial in Retina compared topical difluprednate zero-point-zero-five percent against oral eplerenone and observation in fifty-three patients [9]. Complete subretinal fluid resolution occurred in ninety percent of the difluprednate arm versus forty-five percent with eplerenone and twenty-three percent with observation, with greater reductions in subfoveal choroidal thickness and better microperimetry sensitivity. That is a striking result, and a counterintuitive one given that corticosteroids are usually implicated as a cause of central serous disease rather than a treatment. Note also that thirty percent developed ocular hypertension requiring topical therapy. This is a small, unmasked pilot, and it needs replication before anyone changes practice — but it is a provocative finding worth watching. Rounding out the laser theme, a PROSPERO-registered meta-analysis in Ophthalmology Retina pooled fifteen studies and seven hundred and eighteen eyes treated with Nd:YAG vitreolysis for primary symptomatic floaters, finding an almost nineteen-point improvement in the National Eye Institute Visual Function Questionnaire composite score and improved contrast sensitivity, with unchanged acuity and infrequent, generally transient adverse events [6]. Only two of the fifteen studies were randomized, so certainty is limited, but for carefully selected patients with amenable opacities this supports vitreolysis as a low-risk option short of vitrectomy.

Our final theme is corneal care — one paper on graft quality, one on access. In Ophthalmology, a pre-specified secondary analysis of the Diabetes Endothelial Keratoplasty Study examined what predicts endothelial cell density one year after Descemet membrane endothelial keratoplasty in more than twelve hundred eyes [1]. Beyond the expected relationship with donor preoperative cell density, three factors independently predicted worse outcomes: eye-bank-reported endothelial damage of ten to twenty percent on trypan blue staining after lenticule preparation, which was associated with thirty-six percent cell loss versus twenty-nine percent when damage was under five percent; intraoperative complications, with forty-two percent loss versus twenty-seven percent; and a recipient indication other than Fuchs dystrophy, with forty percent loss versus twenty-eight percent. The practical point is that trypan blue damage assessment after preparation is an independent measure of donor health that complements cell density, and eye banks and surgeons should be using both. And in the American Journal of Ophthalmology, a perspective essay raises an access problem that United States clinicians need to be aware of now [10]. Corneal cross-linking demonstrably slows keratoconus progression and reduces transplant need, but access has been constrained by cost, insurance coverage and proprietary protocols. With the 2025 approval of the epithelium-on product Epioxa under orphan drug designation at a reported wholesale acquisition cost of seventy-eight thousand five hundred dollars, and the planned phase-out of Photrexa in 2026, the United States could be left with no approved epithelium-off cross-linking option and a substantially more expensive alternative of unproven long-term equivalence. The authors call for pausing the phase-out and forming an interdisciplinary access taskforce. This is advocacy rather than data, but it is the kind of structural change that will directly affect which of your patients get treated.

If you only have time for one paper this week, make it the JAMA Ophthalmology cohort on cystoid macular edema after cataract surgery in previously vitrectomized eyes [4]. It's the largest and cleanest dataset here, and it changes a conversation you have in clinic almost every week.

Here are the key takeaways from this week in Ophthalmology. First, a prior vitrectomy is associated with a roughly three-and-a-half to four-fold higher risk of cystoid macular edema after cataract surgery, and that belongs in your consent and your follow-up planning, even though the data cannot tell us whether prophylaxis helps. Second, treating retinal breaks with either laser or cryotherapy is associated with more epiretinal membrane than leaving them alone, with no difference between modalities — likely reflecting vitreoretinal disruption in general, and probably confounded by break severity. Third, in giant retinal tear disease, a third of fellow eyes detached, and a light anterior barrier laser showed no protective effect, so surveillance rather than reassurance is the order of the day. Fourth, in macular disease, high-intensity lipophilic statins showed an association with lower rates of macular degeneration incidence and progression while any-statin use did not — intriguing but not yet actionable; subthreshold laser in acute central serous disease is safer than conventional laser but resolves fluid less reliably; and topical difluprednate in chronic disease outperformed eplerenone and observation in a small pilot, at the cost of ocular hypertension in nearly a third. And fifth, on the cornea side, trypan blue assessment of endothelial damage after lenticule preparation independently predicts one-year graft cell loss, while the impending phase-out of Photrexa threatens to widen existing disparities in cross-linking access across the United States.

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.