Prognosis and prognostic factors of retinal infarction: a prospective cohort study

OBJECTIVE: To determine the prognosis and adverse prognostic factors in patients with retinal infarction due to presumed atheromatous thromboembolism or cardiogenic embolism.

DESIGN: Prospective cohort study.

SETTING: University hospital departments of clinical neurology.

PATIENTS: 99 patients with retinal infarction, without prior stroke, referred to a single neurologist between 1976 and 1986 and evaluated and followed up prospectively until death or the end of 1986 (mean follow up 4.2 years).

INTERVENTIONS: Cerebral angiography (55 patients), aspirin treatment (37), oral anticoagulant treatment (eight), carotid endarterectomy (13), cardiac surgery (six), and peripheral vascular surgery (two).

MAIN OUTCOME MEASURES: Death, stroke, coronary events, contralateral retinal infarction; survival analysis confined to 98 patients with retinal infarction due to presumed artheromatous thromboembolism or cardiogenic embolism (one patient with giant cell arteries excluded), and Cox's proportional hazards regression analysis, including age as a prognostic factor.

RESULTS: During follow up 29 patients died (21 of vascular causes and eight of non-vascular or unknown causes), 10 had a first ever stroke, 19 had a coronary event, and only one developed contralateral retinal infarction. A coronary event accounted for more than half (59%) of the deaths whereas stroke was the cause of only one death (3%). Over the first five years after retinal infarction the actuarial average absolute risk of death was 8% per year; of stroke 2.5% per year (7.4% in the first year); of coronary events 5.3% per year, exceeding that of stroke; and of stroke, myocardial infarction, or vascular death 7.4% per year. Prognostic factors associated with an increased risk of death were increasing age, peripheral vascular disease, cardiomegaly, and carotid bruit. Adverse prognostic factors for serious vascular events were increasing age and carotid bruit for stroke, and increasing age, cardiomegaly, and carotid bruit both for coronary events and for stroke, myocardial infarction, or vascular death.

CONCLUSIONS: Patients who present with retinal infarction due to presumed atherothromboembolism or cardiogenic embolism are at considerable risk of a coronary event. The risk of stroke, although high, is not so great. Not all strokes occurring after retinal infarction relate directly to disease of the ipsilateral carotid system, although this is probably the most common cause. Few patients experience contralateral retinal infarction. Non-arteritic retinal infarction should be diagnosed or confirmed by an ophthalmologist, and the long term care of patients with the condition should involve a physician who has an active interest in managing vascular disease.