Author: Claude (AI), from PubMed abstracts · Reviewed by: Dr Mina Abdelmseih · Reviewed: 2026-11-15 · Review due: 2027-11-15

Educational content only. Not personalised medical advice. Clinical decisions remain the responsibility of the treating clinician.

Central Retinal Artery Occlusion: An Ocular Stroke

Central retinal artery occlusion (CRAO) is a form of acute ischaemic stroke that causes severe visual loss and heralds further cerebrovascular and cardiovascular events [S4]. It is a medical emergency [S4]. No treatment has proven benefit in randomised trials [S6] [S11] [S12], so the priorities are fast triage, excluding giant cell arteritis, and secondary prevention [S4] [S9] [S8].

Key points - Treat CRAO as a stroke: systems of care should prioritise early recognition and emergency triage [S4] [S8]. - Giant cell arteritis (GCA) can cause CRAO; treatment must start urgently when GCA is suspected with eye signs, to prevent blindness [S9] [S1]. - Observational individual participant data suggest better recovery with early thrombolysis [S10] [S7], but two phase 3 RCTs within 4.5 hours found no significant benefit over aspirin [S11] [S12]. - Conservative measures such as eyeball massage, oxygen or pressure lowering have no proven benefit; the evidence is uncertain [S6]. - Every patient needs urgent screening and treatment of vascular risk factors [S4] [S8].

Evidence last checked: 2026-10-17 (PubMed abstracts).

Red flags

• Sudden, severe, persistent loss of vision in one eye, the presentation enrolled in the THEIA trial [S11]. CRAO is a medical emergency [S4].
• Headache, scalp tenderness or jaw claudication, especially in women over fifty: suspect GCA [S9]. GCA is an ophthalmic emergency needing early diagnosis and immediate high-dose steroid to protect both eyes [S1].
• Neurological symptoms with the eye event need neurological evaluation [S2].
• High risk of further vascular events. Patients have very high cardiovascular risk [S8] and CRAO is a harbinger of stroke and cardiovascular events [S4].

Recognise it

• Retinal artery occlusion is associated with increased age and cardiovascular risk factors [S3].
• Embolism is the most common cause [S2].
• Subtypes differ in prognosis. In eyes seen within 7 days with counting fingers or worse, vision improved spontaneously in 82% of transient non-arteritic CRAO, 67% with cilioretinal artery sparing and 22% of non-arteritic CRAO [S1]. Final acuity of 20/40 or better was seen in 89% of permanent branch retinal artery occlusion [S1].

Confirm it

• Record the time of onset: time to treatment influenced visual outcome in observational data [S10] [S7].
• Exclude GCA. Temporal artery biopsy is the gold standard; Doppler ultrasound and vascular imaging are also reliable [S9]. Do not delay treatment for tests when GCA is strongly suspected [S9].
• Find the embolic source. Evaluate the carotid artery and heart, with fasting lipids and a complete blood count [S2].

Manage it

Before acting on this section, check the national and local guidelines, product information and regulatory approvals that apply where you practise. See the medical disclaimer.

Tier A (primary care / no specialist)

1. Treat as an emergency: arrange immediate transfer to emergency care, ideally where stroke expertise is available [S4] [S3].
2. Ask about GCA symptoms [S9]. If GCA is strongly suspected with eye signs, start treatment urgently; the French protocol reports prednisone 40-80 mg per day as initial treatment (as reported in [S9]). Follow local protocols for steroid route and dose.
3. Do not delay transfer for eyeball massage or other conservative measures; none has proven benefit [S6].
4. Start risk-factor control with primary care afterwards [S3] [S4].

Tier B (general eye clinic)

• Confirm the diagnosis and subtype, including cilioretinal sparing and branch occlusion, which carry a better prognosis [S1].
• Screen for GCA and arrange urgent treatment [S9].
• Arrange carotid and cardiac evaluation for embolism [S2] and refer for vascular risk-factor treatment [S4] [S8].
• Coordinate care with neurologists with stroke expertise [S3].

Tier C (specialist centre)

1. Intravenous thrombolysis: conflicting evidence. - In an individual participant data meta-analysis of 783 patients, intravenous thrombolysis within 4.5 hours was associated with recovery from severe vision loss in 28.8% vs 11.1% with conservative therapy (odds ratio 3.32); each hour earlier was associated with better acuity [S10]. Another meta-analysis of 771 patients reported 0.3 logMAR or more improvement in 74.3% after intravenous tPA within 4.5 hours [S7]. Both used non-randomised data. - THEIA (France, 70 patients): alteplase 0.9 mg/kg vs aspirin 300 mg within 4.5 hours (as reported in [S11]) gave improvement in 66% vs 48%, not statistically significant; the trial was likely underpowered and had no symptomatic haemorrhages [S11]. - TenCRAOS (78 patients, six countries): tenecteplase 0.25 mg/kg vs aspirin 300 mg (as reported in [S12]) gave vision recovery in 20% vs 24% at 30 days, with more adverse events including one fatal intracranial haemorrhage [S12]. - The AHA statement (2021, expert consensus) concluded that intravenous tPA may be effective [S4]. Benefit has not been confirmed in RCTs [S11] [S12].
2. Intra-arterial thrombolysis. A meta-analysis of 5 studies found any visual improvement more likely within 6 hours of onset (odds ratio 1.703) but not beyond [S5]. Individual participant data showed a changepoint at 8 hours [S10].
3. Other interventions (hemodilution, pentoxifylline, transcorneal electrical stimulation) showed no clear benefit in small RCTs [S6].

Refer when

Pitfalls

• Treating CRAO as an eye-only problem and missing the stroke risk [S4].
• Missing GCA, which threatens the fellow eye [S1] [S9].
• Overstating thrombolysis: observational data are positive but RCTs are not [S10] [S11] [S12].
• Wasting time on unproven conservative manoeuvres [S6].

Evidence notes

• Conflict: observational meta-analyses favour early thrombolysis [S7] [S10]; the two RCTs did not show benefit and one raised safety concerns [S11] [S12]. A trial would need about 95 participants per group to detect an odds ratio of 3.0 [S10].
• Conflict: one expert review recommends carotid and cardiac evaluation rather than neurological evaluation unless there are neurological symptoms [S2]; other sources favour a multidisciplinary stroke approach [S3] [S4].
• Spontaneous improvement occurs in all subtypes, which complicates uncontrolled studies [S1].
• Not covered by these sources: antiplatelet choice and duration, hyperbaric oxygen, neovascular complications after CRAO, and options where no stroke unit is available.

Related articles

• Giant Cell Arteritis
• Sudden painless vision loss: triage
• Retinal Vein Occlusion

References

See the numbered source list below. Each marker [S#] links to its PubMed and DOI record.