Central retinal artery occlusion and why it is important for neurological health

Central retinal artery occlusion (CRAO) is a devastating and usually permanent cause of blindness.
 
We are working toward evaluating the efficacy of tPA for acute CRAO and our evidence preliminarily suggests this may be beneficial if administered early, within 4.5 hours of the onset of blindness.

We conducted a national survey to evaluate how CRAO is treated in academic centers in the USA, and discovered that practive was highly variable.  Some centers provide numerous treatments and thorough evaluation of stroke risk factors, while others conduct a brief outpatient evaluation only. 

Among our patients, we found that approximately 37% had concurrent acute strokes on MRI (often clinically silent), 37% had critical carotid artery disease and 20% had a critical finding on echocardiogram.  A full 25% of our patients required an acute surgery to address a stroke risk factor. In the year after CRAO, the rate of combined stroke, myocardial infarction and death was 25%, indicating the high rate of concurrent vascular risk in the CRAO population. 
Central retinal artery occlusion is an ophthalmologic emergency and an important cause of acquired blindness.  Occlusion of the central retinal artery is usually secondary to one or more serious systemic diseases, often carotid artery or cardiac valvular disease, although hypercoagulability, atrial fibrillation and autoimmune diseases are significant contributors as well.  While most patients with acute occlusion often present with minimal or no ability to detect light, a minority of patients have collateral supply to the retina via a cilioretinal artery and may present with less severe visual acuity impairment, although they often still have severe functional impairment.  The central retinal artery narrows as it pierces the dura surrounding the optic nerve at its midpoint and again as it passes through the lamina cribrosa (a few millimeters before it reaches the retina) – emboli usually lodge at one of these points, and for this reason in most patients the embolus cannot be visualized on funduscopy.  The occlusion is inferred on funduscopic examination based on narrowing of the retinal vessels, pallor of the retina producing a red macula (the so-called cherry red spot), occasionally by optic nerve head swelling, and visualization of distal microemboli in the branch retinal arteries.  In the acute setting, the fundus may appear normal—the only objective finding being a relative afferent pupillary defect; although fluorescein angiography or retinal Doppler sonography could confirm the diagnosis in this setting.  A minority of patients (approximately 17%) have improvement in visual acuity without any treatment, presumably due to spontaneous reperfusion of the retina prior to the onset of permanent damage.  Usually, however, patients develop permanent blindness.  Despite being a topic of research for nearly 150 years, there is still no proven treatment to restore visual acuity outcomes and no consensus for how these patients should be evaluated and treated.  As a result, treatment approaches have evolved based on local experience and/or beliefs regarding the pathophysiology of this disease.
Medial to lateral view of the vascular supply to the optic nerve and retina

The order of the major branches off the ophthalmic artery is variable; the most common configuration is shown at left.  The posterior segment of the optic nerve is supplied by the pial plexus primarily via branches from the ophthalmic artery.  The anterior segment of the optic nerve is supplied primarily by branches from the posterior ciliary arteries.  The retina is primarily supplied by the central retinal artery and in a minority of cases has additional supply from the cilio-retinal artery, a branch of the posterior ciliary artery.  Collateral arterial blood flow from the meningeal circulation is variable and in cases of chronic vascular disease may be the dominant arterial supply to the central retinal and posterior choroidal arteries.
Trends in CRAO treatment in the USA:
We identified university-associated teaching hospitals offering vascular neurology, neuro-ophthalmology and/or retina fellowships in the United States and asked the directors of the programs to respond to questions in an open response format to profile the acute management of CRAO at their institution. 

We found remarkable heterogeneity in the approach to acute treatment of patients with CRAO among the 45 institutions that responded to the survey.  Only 20% had a formal policy, guideline or white paper to standardize the approach to treatment.  The primary treating physician was an ophthalmologist, neurologist, or neuro-ophthalmologist 44, 27, and 7% of the time, respectively; 24% were co-managed acutely by neurology and ophthalmology.  Intravenous fibrinolysis was offered to selected patients in 53% of institutions, and was the preferred initial treatment in 36%; anterior chamber paracentesis, ocular massage and hyperbaric oxygen therapy were offered occasionally, while 9% of institutions offered no treatment.  At 35% of institutions, patients with acute CRAO were not referred to a general emergency room for initial treatment routinely, and overall the approach to risk modification was highly variable. 

Currently no consensus exists on how patients with acute CRAO ought to be managed, leading to widespread variability in treatment practices.  Due to the high rate of comorbid disease, it is essential that these patients be admitted for emergent inpatient evaluation, ideally with co-management by ophthalmology and vascular neurology.  We present an evidence-based protocol for the evaluation and management of acute CRAO.
Central retinal artery occlusion carries the same risk profile as a stroke or TIA, so it should be treated as a stroke/TIA equivalent.
Patients with central retinal artery occlusion need immediate evaluation and management of their cardiovascular and cerebrovascular risk factors.

tPA for acute CRAO

... and lack of efficacy of anterior chamber paracentesis, hemodilution and ocular massage


Central retinal artery occlusion  (CRAO) is a relatively rare disorder that is caused by interruption of blood flow to the retina, usually by a clot or some other embolus.  Despite around 150 years of research, no compelling treatment has been found for this disease.  Treatment with fibrinolytics has been used experimentally for a long time and some of the results have been encouraging.  We conducted a patient-level meta-analysis to aggregate all of this observational data and compare how patients with CRAO do when treated with fibrinolytics versus when they are treated with other approaches or not treated at all.  This type of analysis tries to maximize the usefulness of the best retrospective and observational data in aggregate to help guide clinical decision making and contribute to better design of clinical trials to hopefully improve the odds of having a successful trial. 
 
The biggest surprise in the data was the poor performance of conventional treatments at less than half the recovery rate of patients who were simply left alone.  While this result is limited by the retrospective, non-randomized study design, it raises enough doubt that I think ocular massage, anterior chamber paracentesis and hemodilution should be abandoned as treatments for acute CRAO.  The data is encouraging regarding the effectiveness of tPA for CRAO -- but only in the first 4.5 hours from symptom onset.  This is not strong enough data to make a compelling recommendation that patients with CRAO should receive tPA.  However, many centers do treat CRAO off-label with tPA and for those centers I would recommend only treating within the first 4.5 hours. 

Full text ....
here.
Forest plot for estimated rate of spontaneous visual recovery after CRAO.  (Recovery defined as initial visual acuity of counting fingers or less, improving to better than 20/200.)Subjects in natural history cohorts (those with no or minimal treatment) had a spontaneous recovery rate of 17.7% (95% CI, 13.9-21.4%).  Those treated with “conservative” treatment including ocular massage, anterior chamber paracentesis and/or hemodilution had a significantly lower recovery rate of 7.4%, (95% CI, 3.7-11.1%).  Subjects treated with intravenous fibrinolysis had recovery in 31.8% (95% CI, 24.3-39.3%).  There is heterogeneity only in the “conservative treatment” group, due to the impact of a single outlier study (exclusion of this study would reduce measurements of heterogeneity to non-significant levels). 

tPA treatment within 4.5 hours resulted in a significantly higher rate of visual recovery compared to the natural history cohort; 50.0% recovery, 95%CI, 32.4-67.6%,  vs. 17.7%, 95%CI, 13.9-21.4% (p<0.0001).  There was no statistical benefit to treatment more than 4.5 hours after onset.