麻豆传媒

Patients with retinal artery occlusions (RAOs) had subsequent increased risks of death, stroke, and myocardial infarction (MI) over both the short and long term compared with matched controls with cataracts, a retrospective cohort study showed.

In a propensity score-matched analysis using aggregated electronic health records (EHRs), the relative risk (RR) of death after RAO was significantly higher compared with control patients at all time points measured:

• 2 weeks: RR 2.45, 95% CI 1.46-4.12
• 30 days: RR 2.10, 95% CI 1.49-2.97
• 1 year: RR 1.78, 95% CI 1.61-1.94
• 5 years: RR 1.28, 95% CI 1.23-1.33
• 10 years: RR 1.05, 95% CI 1.02-1.07

Stroke after RAO compared with matched controls followed a similar pattern: the relative risk for stroke was approximately 21-fold higher in the first 2 weeks, 14-fold higher at 30 days, and five-fold higher within 1 year, reported Prithvi Mruthyunjaya, MD, MHS, of Byers Eye Institute and Stanford University School of Medicine in Palo Alto, California, and colleagues in .

MI risk was considerably lower, but followed a similar trajectory, declining from three-fold higher in the first 2 weeks, to 2.6-fold higher within 30 days, and 1.7-fold higher within 1 year.

With no proven therapy to address vision loss due to RAO, management is aimed at averting further complications of RAO and "preventing future secondary vascular events with cardiovascular evaluation and optimization of ," the authors wrote.

Robin A. Vora, MD, of Kaiser Permanente Northern California, told 麻豆传媒 that central RAO is known to be a vision-threatening disease, and this study shows that it "may additionally indicate underlying severe cardiovascular and/or neuro-vascular disease."

"As such, it is essential that the diagnosing ophthalmologist refer these patients for urgent evaluation that may include neuro- and neuro-vascular imaging, carotid Dopplers, electrocardiography, echocardiography, and event monitoring," he said. "Rapid referral to a stroke center is sensible as any identified risk factors can be more quickly addressed, lowering risk of an adverse event subsequent to diagnosis of central RAO."

Indeed, Mruthyunjaya and co-authors noted that mixed reports regarding the strength of the association between RAO and secondary vascular ischemic events have led to variations in post-RAO management. They pointed to a that showed that "within 12 hours after diagnosis of RAO, 75% of neurologists pursue a hospital-based evaluation, whereas 82% of retina specialists pursue an outpatient workup."

Importantly, they added, both the and the recommend that all patients with RAO receive an emergency cardiovascular workup.

Vora said that "one important point not mentioned in the article is that physicians should always consider the possibility of giant cell arteritis (GCA) as the initial cause of central RAO, [as] GCA is a serious systemic condition that must not be missed."

"In our center, patients with acute central RAO (<48 hours of symptoms) are referred to the emergency room for rapid evaluation and testing," he noted. "We have collaborated with our stroke neurologists and emergency room physicians to ensure these patients receive the same prompt management as would patients presenting with TIA [transient ischemic attack] or stroke."

For this study, Mruthyunjaya and team used aggregated electronic health record data from January 2003 through April 14, 2023 from TriNetX, which includes data on more than 111 million patients, to identify patients with an ICD-10 diagnosis code for RAO or age-related cataract. Those who had experienced a stroke or MI within 2 years before RAO or cataract diagnosis were excluded.

They included 34,874 patients with at least 1 year of follow-up in the RAO cohort. Mean age at RAO event was 66, 51% were women, and 71% were white. Of these patients, 24% had central RAO, 35% had transient RAO, and 42% had "other" RAO.

RAO patients were matched for age, sex, race, and comorbidities, including hypertension (55%), diabetes (26%), hyperlipidemia (49%), and smoking status (12% had nicotine dependence), to 34,552 control patients.

Mruthyunjaya and colleagues noted that study limitations included those inherent in the use of aggregated electronic health record data, and reliance on accurate ICD-10 diagnosis coding.

Comment