Patients who have previously undergone pars plana vitrectomy (PPV) may face a greater risk of developing cystoid macular edema (CME) after cataract surgery, according to a retrospective cohort study published in JAMA Ophthalmology.
Charles Zhang and colleagues at Bascom Palmer Eye Institute, Florida, analyzed electronic health records from the TriNetX US Network, encompassing academic and community hospitals between December 2005 and December 2025. The study included 615,983 adults who underwent cataract surgery, of whom 7,422 had undergone PPV more than six months previously.
To minimize confounding, the researchers excluded patients with pre-existing CME or retinal edema, uveitis, retinal vein occlusion, exudative age-related macular degeneration, diabetic retinopathy, or a history of intravitreal injections. Propensity score matching was then used to balance demographic and clinical characteristics, producing 7,318 matched pairs.
Within 30–90 days of cataract surgery, CME was recorded in 336 patients with previous PPV (4.59%), compared with 90 patients without prior PPV (1.23%), representing an absolute difference of 3.36 percentage points and a relative risk of 3.73.
The association remained evident when patients were stratified according to the indication for vitrectomy. Among those whose PPV was related to retinal detachment, CME occurred in 5.65 percent of patients, versus 1.22%of matched controls. For PPV performed for non-retinal-detachment indications – including vitreous hemorrhage, epiretinal membrane, macular hole, and vitreous opacities – the corresponding rates were 3.99 and 1.23%.
The increased risk also persisted after excluding patients with coded intraoperative or postoperative complications of cataract surgery: CME occurred in 4.59%of the prior-PPV group and 1.26%of controls.
One possible explanation is that removing the vitreous eliminates a barrier to the posterior diffusion of inflammatory mediators released during cataract surgery. Vitrectomy may also produce longer-term alterations in intraocular oxygenation and inflammatory signaling. However, the study authors acknowledged that closer retinal follow-up and more frequent OCT imaging in previously vitrectomized patients could have increased CME detection, leading to potential surveillance bias for the observed association.
Other limitations include reliance on diagnostic and procedural codes, absence of laterality data, potential residual confounding, and a lack of visual-acuity outcomes. Consequently, the study cannot establish causality or determine whether the additional CME diagnoses resulted in clinically meaningful visual loss.
The findings may support more careful counseling and postoperative surveillance of vitrectomized eyes, but the authors stopped short of recommending a specific prophylactic regimen. Prospective studies will be needed to establish whether intensified anti-inflammatory treatment or altered follow-up can improve outcomes in this population.
