Audio link: 

https://open.spotify.com/episode/2qzQXzcOdUUfz4N6Bf3sjI?si=eAaqyucbSSS_GNCU1a_OcA

In this week’s issue

• Myopic eyes with deformities on 3D-MRI had a higher risk of myopic complications over 15 years compared to nondeformed myopic eyes.
• Former smokers who switched to electronic cigarettes or heated tobacco had a higher risk of developing eye diseases compared to those who quit completely.
• In highly myopic children, an AI model accurately predicted future pathologic myopia based on early microvascular and structural retinal changes.
• A greater baseline choroidal thickness predicted better long-term visual acuity in patients with intermediate uveitis.
• 
  

Ophthalmology

Toward an era of oculomics: association between baseline diabetic retinopathy severity and mortality risk

Diabetic retinopathy severity correlates with rates of all-cause mortality. Diabetic retinopathy (DR) is well recognized as a marker of systemic microvascular disease, but its relationship with long-term mortality is incompletely characterized. In this retrospective cohort study, researchers linked electronic health record data from over 524,000 patients with diabetes mellitus (DM) to National Death Index records to assess whether baseline DR severity predicts mortality. Patients were categorized at their first eye care visit into six groups based on DR severity in the more affected eye: no DR, nonproliferative DR (NPDR) without or with diabetic macular edema (DME), proliferative DR (PDR) without or with DME, and unspecified DR. Cox regression modeling was used to estimate the hazard of death associated with each category, with adjustment for clinical variables including HbA1c, as well as nonclinical factors such as age, sex, and socioeconomic status. Over a mean follow-up of 5.4 years, 7% of patients died. After adjustment, NPDR was associated with a 31–34% higher mortality risk versus DM without DR, while PDR was associated with a substantially higher risk (128% greater for PDR without DME and 87% greater for PDR with DME). Older age, male sex, and lower neighborhood affluence were also independently associated with increased mortality. These findings reinforce that DR severity is not merely an ophthalmic concern, but a powerful systemic prognostic indicator, positioning the eye exam as a valuable opportunity to flag patients who may benefit from more intensive medical management of their systemic disease.

JAMA Ophthalmology 

Eye shape as a prognostic marker in high myopia

The shape of things might make a difference. Pathologic myopia is defined by structural complications in the fundus. Myopia has traditionally been characterized by axial length and refractive error, which do not adequately capture heterogeneity in disease progression. 3D-MRI offers information on ocular geometry. This study used 3D-MRI to track eye shape and outcomes in 152 highly myopic eyes (78 participants) followed for 15 years in the Zhongshan High Myopia Cohort. At baseline, eyes were classified as nondeformed (spheroidal and ellipsoidal, 61.8%) or deformed (nasally distorted, conical, temporally distorted, and barrel-shaped, 38.2%). Deformed eyes elongated significantly faster than nondeformed eyes (0.082 vs 0.052 mm/year) and had a higher risk of choroidal thinning, myopic macular degeneration (MMD), and visual field defects. Nasally distorted eyes had the highest-risk phenotype overall, with the fastest elongation and greatest risk of MMD progression (OR 4.75), visual impairment (OR 7.40), and visual field defects (OR 14.98). Conical eyes showed comparable risk for choroidal thinning and field loss. In contrast, temporally distorted and ellipsoidal eyes had relatively favorable long-term outcomes despite continued elongation and shape abnormality. These results show that morphology predicts long-term risk in high myopia, offering a new approach to risk stratification and targeted surveillance, particularly for nasally distorted and conical eyes.

American Journal of Ophthalmology 

Quitting smoking versus switching to non-combustible nicotine products... Are the eyes still paying the price?

The smoke may be gone, but the ocular consequences of nicotine exposure may still linger. Non-combustible nicotine or tobacco products (NNTPs), including electronic cigarettes and heated tobacco, are becoming increasingly popular among former combustible cigarette smokers. Recent studies have linked electronic cigarette use to ocular symptoms and tear-film abnormalities, and smokeless tobacco use to late age-related macular degeneration. However, whether switching from combustible cigarettes to NNTPs is risk-neutral compared to complete smoking cessation for eye health remains unclear. To address this, investigators conducted a nationwide retrospective cohort study of 179,273 former smokers in Korea. After propensity score matching, two cohorts of 16,158 participants each (NNTP users and complete quitters) were followed for up to 6 years. NNTP switchers had a 7% higher risk of developing major vision-impairing eye diseases compared with complete quitters. The strongest associations were observed for diabetic retinopathy and refractive and accommodation disorders. No significant associations were found for cataract, glaucoma, or AMD individually. Limitations of the study include self-reported nicotine exposure and a relatively limited follow-up for slowly progressive ocular diseases. Nevertheless, NNTP use after smoking cessation appears to sustain residual risk of vision-impairing eye disease, likely through ongoing nicotine-related microvascular and inflammatory pathways. These findings support complete cessation of all tobacco and nicotine products as the preferred strategy for preserving vision.

British Journal of Ophthalmology

Can AI predict which children with high myopia will develop pathologic myopia?

AI may be helping us see the future of myopia—literally. High myopia (HM) in children and adolescents is becoming increasingly prevalent worldwide, raising concern for future vision-threatening complications such as pathologic myopia (PM). However, most existing research has focused on the late-stage manifestations of PM, primarily in adults, leaving early predictive markers in pediatric populations not fully understood. In this prospective longitudinal study, investigators aimed to identify biomarkers predictive of PM in 375 children with HM (spherical equivalent ≤ −6.00 diopters) by analyzing retinal vasculature from fundus photographs using deep learning and integrating longitudinal ocular biometry and optical coherence tomography metrics. Researchers developed an artificial intelligence (AI)-based model to estimate future PM risk. After multivariable adjustment, independent predictors of PM included lower retinal vessel density, reduced fractal dimension, narrower arteriolar and venular calibers, thinner subfoveal choroid, and faster early axial elongation, suggesting early microvascular compromise and structural retinal changes that precede overt disease. The AI-based model demonstrated excellent predictive performance, with an area under the receiver operating characteristic curve (AUROC) of 0.96, improving to 0.98 when biometric variables were incorporated. These findings suggest that early microvascular and structural retinal alterations may serve as meaningful biomarkers of PM progression and highlight the potential of AI-assisted screening to identify high-risk pediatric patients before irreversible vision loss occurs.

Journal: Scientific Reports

Can the choroid see the future in intermediate uveitis?

Sometimes the most valuable biomarker isn't the one that changes, it's the one that helps predict what comes next. Intermediate uveitis is a chronic inflammatory eye disease that can lead to vision-threatening complications, such as macular edema. Because monitoring often relies on subjective clinical judgment or invasive imaging, noninvasive structural biomarkers are highly sought after. To address this, researchers conducted a retrospective longitudinal study of 91 eyes from 52 patients to evaluate whether subfoveal choroidal thickness (SFCT) and mean choroidal thickness (MCT) on optical coherence tomography (OCT) could serve as reliable prognostic indicators. Multivariable regression models demonstrated that while choroidal thickness remains stable over time regardless of whether a patient clinically worsens, improves, or stays stable, greater baseline SFCT and MCT are both significantly prognostic of better future visual acuity. Furthermore, higher baseline MCT is uniquely associated with a thinner future central retinal thickness, indicating a potential protective effect against chronic macular edema. Although limited by its retrospective design and small clinical subgroups, this study establishes that readily accessible OCT choroidal measurements serve as a cost-effective prognostic biomarker. These findings suggest that preserving choroidal structure may safeguard long-term vision, laying an excellent foundation for future prospective trials.

Journal: Ophthalmology Retina

Atopic dermatitis as a potential risk factor for vision-threatening retinal disease

Looking beyond the anterior segment. Atopic dermatitis (AD) is associated with anterior segment pathology, including keratoconus, cataracts, and ocular surface disease, but its association with retinal pathology has received little attention. A recent retrospective cohort study examined whether patients with AD are at increased risk for retinal detachment (RD) and postoperative proliferative vitreoretinopathy (PVR). Using a large multicenter database, investigators propensity-score-matched 274,161 patients with AD to an equal number of controls and followed both cohorts for 5 years. Patients with AD were more than three times as likely to develop RD (0.7% vs. 0.2%) and more than four times as likely to undergo RD repair. Among patients who underwent surgery, postoperative PVR occurred in 5.9% of AD patients compared with 4.0% of controls. These findings suggest that AD may be associated with an increased risk of both retinal detachment and postoperative PVR, and, if confirmed in future studies, a positive AD history could have implications for counseling and risk assessment.

Scientific Reports

Early detection of keratoconus with AI

Excited to see corneal ectasia earlier! Keratoconus is a progressive corneal disorder characterized by stromal thinning and anterior protrusion. Early detection is critical to preventing irreversible visual loss. A recent study evaluated multiple machine learning algorithms for their ability to classify keratoconus cases correctly. The models were trained on Pentacam corneal topography and biomechanical parameters from 144 corneal scans in adults aged 18 to 35. The models reached 98% accuracy during training and 96% during testing. Deep learning approaches demonstrate AUROCs of 0.96 to 0.97 in pooled analyses, outperforming conventional topographic screening tools. A key contribution of the study is its interpretability. Unlike prior approaches that treat classification as a black box, the authors directly compared features the models selected against clinically recognized diagnostic parameters, identifying which Pentacam outputs most meaningfully differentiated ectatic from normal corneas. These findings provide additional diagnostic value beyond topographic indices alone, and indicate that machine learning can promote early detection of keratoconus.

Don’t get it MUTTled, which antifungal treatment should you use for fungal keratitis? In the 2013 MUTT 1 study, patients with filamentous fungal keratitis were randomized to treatment with natamycin (n = 162) and voriconazole (n = 161). 

Key Points:

• Patients treated with natamycin had significantly better visual acuity measured at 3 months than patients treated with voriconazole.
• Patients with ulcers treated with natamycin were less likely to undergo perforation or transplantation.

Overall, the MUTT 1 study is a landmark study because it showed that natamycin was not dangerous in the treatment of filamentous fungal keratitis and actually led to improved clinical and microbiological outcomes.