Thrombolytic Therapy for Central Retinal Artery Occlusion in an Academic Multi-Site Stroke Centre

ABSTRACT

Central retinal artery occlusion (CRAO) is a subtype of acute ischaemic stroke leading to severe visual loss. A recent American Heart Association scientific statement proposed time-windows for thrombolysis in CRAO similar to acute ischaemic cerebral strokes. We aimed to review our academic multi-site stroke centre experience with intravenous (IVT) and intra-arterial thrombolysis (IAT) in CRAO between 1997 and 2022. Demographic, clinical characteristics, thrombolysis timeline, concurrent therapies, complications, and 3-month follow-up visual acuity (VA) were collected. The thrombolysed cohort follow-up VA was compared with an age, gender and baseline VA matched cohort of CRAO patients that received conservative therapies. Thrombolytic therapy was administered to 3.55% (n = 20) of CRAO admissions; 13 IVT (mean age 68, 61.5% male, 12 alteplase and 1 tenecteplase, all embolic aetiology, 1 CRAO mimic) and 7 IAT (mean age 55, 85.7% male, 3 post-operative and 3 embolic). Additional conservative CRAO-targeting therapies was received by 60%. The median time from onset of visual loss to IVT was 158 minutes (range 67–260). Improvement by at least two Snellen lines was achieved by 25% with 12.5% improving to 20/100 or better. Intracranial haemorrhage post IVT occurred in 1/13 (7.6%). The median time from onset of visual loss to IAT was 335 minutes. Improvement by at least two Snellen lines was achieved by 42%. No difference in 3-month VA was noted between patients that received thrombolysis, either alone (n = 8) or combined with other therapies, and those that received conservative therapies. Our results suggest that the management of acute CRAO remains heterogeneous. The lack of obvious benefit of thrombolysis in our small series supports the need for randomizsd clinical trials comparing thrombolysis to placebo to guide hyperacute CRAO management.

Introduction

Central retinal artery occlusion (CRAO) is a neuro-ophthalmological emergency that can lead to irreversible vision loss¹ with an incidence of approximately 1.8–2.5 per 100 000 person-years.² Acute CRAO is analogous to a brain acute ischaemic stroke (AIS), as the obstruction of the central retinal artery (CRA) causes end-organ ischaemia followed by cell death of the inner retina.^3,4 The CRA is a single 160 μm diameter end artery supplying inner retinal cells that are extremely susceptible to ischaemia due to high numbers of mitochondria, and have limited collateral circulation in case of complete CRA occlusion. The most common aetiology of CRAO is non-arteritic (NA-CRAO), often attributed to thromboembolism, whereas inflammatory or arteritic (A-CRAO) from vasculitic processes, such as giant cell arteritis, are rarer.⁵ Multiple conservative therapies for NA-CRAO (e.g. digital ocular massage, acetazolamide, mannitol, topical, intraocular pressure-lowering agents, corticosteroids, hyperbaric oxygen, anterior chamber paracentesis) are often tried in the setting of an acute CRAO.⁶ However, currently there are no standard practice guidelines for treating NA-CRAO.⁷ The American Academy of Ophthalmology preferred practice patterns recognises that conservative therapies are not effective when compared to the natural history of the disease, and some may even be harmful.^8,9

Although thrombolysis has proven benefit for AIS, this treatment remains under investigation for CRAO. Most CRA emboli are cholesterol or calcific plaques, that are theoretically less amenable to dissolution by the tissue plasminogen activator (tPA). Intravenous (IVT) and intra-arterial (IAT) thrombolysis are also more difficult to study in CRAO due to its relative rarity and the infrequent presentation of NA-CRAO patients to stroke centres in the hyperacute, thrombolysis eligible phase. Patients with AIS are candidates for IVT if they present within 4.5 hours of last known normal.¹⁰ Similarly, meta-analyses of retrospective studies have demonstrated improved visual outcomes in CRAO patients treated with IVT within 4.5 hours of visual loss.^9,11 Mechanical thrombectomy extends the time frame for intervention to 24 hours in certain subsets of AIS patients.¹² Given the small CRA diameter,¹³ mechanical thrombectomy is not considered feasible in CRAO. However, IAT can be performed via cannulation through the femoral artery, subsequent catheterisation through the internal carotid artery to the level of ophthalmic artery, and administration of thrombolysis via the proximal ophthalmic artery.^4,14 Thus, IAT is a studied alternative with a broader treatment window.¹⁵ Current evidence for early thrombolysis in CRAO patients is based on multiple patient-level retrospective studies,^9,11,16–20 and is supported by a recent scientific statement from the American Heart Association/American Stroke Association.² However, in the absence of level 1 evidence data showing net benefit of IVT and IAT in acute NA-CRAO, the use of thrombolysis in routine clinical practice remains a topic of debate. To inform practice, we reviewed all NA-CRAO patients that received IVT and/or IAT at our academic multi-centre stroke programme and appraised their clinical characteristics and visual outcomes, as well as the safety of this therapeutic approach.

Methods

We conducted a retrospective review of all electronic medical records of NA-CRAO patients that were treated with IVT and/or IAT between 1997 and 2022 at any of the Mayo Clinic stroke programmes (Arizona, Minnesota, Florida, and Mid-West Mayo Clinic Health System). This time frame was chosen because alteplase received FDA-approval for the treatment of AIS in 1996.²¹ The study was approved by the Mayo Clinic Institutional Review Board. Informed consent was unnecessary due to the retrospective nature of the study and the use of de-identified clinical data. We included patients that presented to our sites’ emergency departments (ED) with CRAO or developed CRAO during inpatient or peri-operative/peri-procedural stay.

For the CRAO subjects who underwent thrombolysis, we collected baseline characteristics, including demographics (age, sex) and vascular risk factors (hypertension, hyperlipidaemia, diabetes mellitus, prior stroke or transient ischaemic attack [TIA], migraine with aura, coronary artery disease, atrial fibrillation, carotid disease, and prior or current tobacco use). Treatment-related data included: duration of time between last seen normal/vision loss and ED presentation; duration of time between symptom onset and thrombolysis; the route of thrombolysis administration (IVT or IAT); the thrombolytic agent administered and its dose; additional CRAO-targeting conservative therapies; and thrombolysis-related complications. Other variables of interest were CRAO aetiology, length of hospital stay, and mortality.

The presence or absence of ophthalmological examination findings at the initial presentation, such as a cherry red spot, retinal whitening, retinal emboli, box-carring, optic disc pallor/oedema, and arterial narrowing, were recorded. Best corrected visual acuity (VA) was recorded at the initial presentation and at follow-up, that was defined as at least 3 months since the initial presentation. Leveraging the Rochester Epidemiology cohort (REP),²² we identified a control group of age, gender, and baseline VA-matched NA-CRAO patients that received conservative CRAO treatment only. For these control subjects, we recorded treatment type, hospital complications, length of hospital stay, VA at baseline and at least 3 months’ follow-up.

Statistical analysis

Descriptive statistics were used to analyse variables of interest. Chi-square and Fisher’s exact tests were used for comparison of categorical values, as appropriate. A comparative analysis of the follow-up VA for the two matched cohorts was also conducted. Snellen VA was converted to logMAR (logarithm of the minimum angle of resolution) for statistical evaluation using the equation: logMAR= -LOG 10 (fraction of Snellen scale). Based on Lange et al.²³ logMAR equivalencies for low VA were approximated as follows: no light perception (NLP) = 3.0, light perception = 2.7, hand motion = 2.28, and count fingers (CF) = 1.85. A two-tailed p-value less than 0.05 was considered statistically significant.

Results

Patient characteristics

Between 1997 and 2022, a total of 563 patients with CRAO were admitted to one of our institutions. Of these, 3.55% (n = 20) received thrombolysis, with 13 patients receiving IVT and 7 patients receiving IAT. Table 1 presents the patients’ demographic, clinical, and treatment characteristics. The majority (60%, n = 12) of the thrombolysed patients received additional conservative CRAO-targeting therapies (7/13 in the IVT and 5/7 in the IAT cohort). Twelve patients (60%) presented directly to the ED with acute visual loss and were evaluated by the neurology and/or ophthalmology teams. A standard acute stroke evaluation was conducted (including head computed tomography to rule-out intracranial haemorrhage), with the addition of ruling-out arteritic CRAO from giant cell arteritis by clinical examination and laboratory testing. Five patients (25%) were initially seen in an ophthalmology clinic and were subsequently referred to the ED for emergent thrombolysis. Three patients (15%) developed CRAO peri-operatively (one post-internal maxillary artery embolisation for maxillary haemangioma, one post-endoscopic sinus surgery for resection of bilateral olfactory cleft masses and one post-embolisation of a facial arterio-venous malformation). One patient (5%) developed CRAO whilst admitted for an unrelated medical reason (alcohol intoxication).

Table 1.

Demographic and clinical characteristics of the central retinal artery occlusion cohort who received thrombolysis.

Patients characteristics	Intravenous thrombolysis	Intra-arterial thrombolysis
Number of patients	13	7
CRAO mimic (n, %)	1 (7.7%)	0
Mean age (years)	68.58	55.29
Sex, female number (%)	5 (38.5%)	1 (14.3%)
Last seen normal to ED presentation mean (minutes) Range (minutes) Median (minutes)	85.4 31–169 67.5	508.75 100–1050 442.5
Last seen normal/visual loss to thrombolysis mean (minutes) (Range) (minutes) Median (minutes)	160.42 67–260 158	516.14 131–1230 335
Conservative treatment, number (%)	6 (50%)	5 (71.43%)
Hospital/thrombolysis complications, number (%)	1 (8.3%)	1 (14.29%)
Arterial hypertension, number (%)	11 (91.67%)	3 (42.86%)
Hyperlipidaemia, number (%)	12 (100%)	5 (71.43%)
Migraine history, number (%)	1 (8.3%)	0
Stroke/TIA history, number (%)	5 (41.67%)	2 (28.57%)
CAD/MI history, number (%)	5 (41.67%)	3 (42.86%)
Atrial fibrillation, number (%)	0 (0%)	2 (28.57%)
Diabetes mellitus, number (%)	1 (8.3%)	4 (57.14%)
Carotid disease, number (%)	4 (33.33%)	2 (28.57%)
CRAO aetiology Thrombo-embolic Cardio-embolic Perioperative Unknown	11 (91.66%) 1 (8.33%)	3 (42.86%) 3 (42.86%) 1 (14.29%)
ESR and/or CRP High Normal N/A	3 (25%) 9 (75%)	3 (42.86%) 4 (57.14%)
Temporal artery biopsy, number (%)	2 (16.67%)	0
Neurology consults alone, number (%)	5 (38.46%)	3 (42.86%)
Ophthalmology consults without neurology, number (%)	0	1 (14.29%)
Both neurology and ophthalmology consults, number (%)	8 (61.53%)	3 (42.86%)
Neuroimaging CT scan without contrast CT angiography head and neck MRI brain	12 (100%) 10 (83.33%) 7 (58.33%)	6 (85.71%) 2 (28.57%) 5 (71.43%)
CRAO findings: Retinal whitening Cherry red spot Retinal emboli Optic disc pallor Optic disc edema Arterial narrowing Arterial box-carring No available data	8 (66.67%) 7 (58.33%) 1 (8.3%) 1 (8.3%) 2 (16.67%) 3 (25%) 4 (33.33%) 1 (8.3%)	3 (42.86%) 4 (57.14%) 1 (14.29%) 0 2 (28.57%) 4 (57.14%) 0 0
Antithrombotic at discharge: Aspirin 81 mg only Aspirin 81 mg and clopidogrel 75 mg Aspirin 325 mg only Aspirin 325 mg and clopidogrel 75 mg Clopidogrel 75 mg and warfarin Ticagrelor 90 mg and aspirin 81 mg Clopidogrel 75 mg and apixaban 5 mg Aspirin/dipyridamole (Aggrenox) None	3 (25%) 4 (33.33%) 2 (16.67%) 1 (8.3%) 1 (8.3%)	1 (14.29%) 2 (28.57%) 1 (14.29%) 1 (14.29%) 1 (14.29%) 1(14.29%)
Length of hospital stay mean (range) (days)	2.63 (1–5)*	2.86 (1–7)
Baseline VA, logMAR mean (± SD)	2.375 (0.54)	2.614 (0.414)
Follow-up VA logMAR mean (± SD) (number)	1.985 (1.00)	1.986 (0.73)

*Excluding one outliner that was admitted for 12 days due to polysubstance abuse.

CAD/MI: coronary artery disease/myocardial infarction; CRAO = central retinal artery occlusion; CRP = C-reactive protein; CT, = computed tomography; ESR = erythrocyte sedimentation rate; logMAR = logarithm of the minimum angle of resolution; MRI = magnetic resonance imaging. TIA = transient ischaemic attack; VA = visual acuity.

IVT cohort

Thirteen patients (65% of the cohort that received thrombolysis) presented within the therapeutic window for IVT (defined as within 4.5 hours of last known normal) and received IVT once contraindications were ruled-out and informed consent was obtained. The mean age of this subset was 68.5 ± 8.2 years (range 55–82) and 61.5% were male. The median time between last seen normal to ED presentation was 67.5 minutes (range 31–169). Twelve patients received alteplase (0.9 mg/kg, 10% administered as a bolus and the remainder infused over 1 hour, according to the institutional acute stroke protocol), and one received tenecteplase (0.25 mg/kg administered as a single bolus). One patient from the IVT group was ultimately categorised as a CRAO mimic, where a nuclear cataract was determined to be the cause of visual loss after subsequent examination by the ophthalmologist; this patient was excluded from further comparative analyses. Vascular neurology was consulted in all cases, whereas ophthalmology was consulted before IVT administration in 61% of cases. One patient received tPA at an outside institution, as recommended by the vascular neurology team. The median time from onset of visual loss to IVT administration was 158 minutes (range 67–260). One patient (7.6%) developed severe symptomatic intracerebral haemorrhage (sICH).

CRAO aetiology was thromboembolic in all patients that received IVT, with moderate or severe carotid stenosis (defined as greater than 50%) in 38.4% of cases (n = 5). One case had a cardio-embolic aetiology. The median length of stay was 3 days (range 1–5), excluding one outlier that was admitted for 12 days due to polysubstance use disorder. All patients were discharged home except for one patient who was discharged to a hospice. Three patients (23.0%) underwent carotid endarterectomy (CEA) and one patient (7.7%) underwent carotid stenting and angioplasty.

IAT cohort

Seven patients (35% of the cohort that received thrombolysis) received IAT, 85.7% of whom were males with a mean age of 55.2 ± 14.2 years (range 17–83). The dose of intra-arterial alteplase ranged from 5 to 30 mg. The median time between last seen normal and ED presentation was 442.5 minutes (range 100–1050). Vascular neurology was consulted in 85.7% of cases and ophthalmology was consulted before IAT administration in 57.1% of cases. CRAO aetiology was thromboembolic in 42.8% of cases (n = 3), peri-operative in 42.8% (n = 3), and unknown in 14.2% of cases (n = 1). The median time from set of visual loss to IAT administration was 335 minutes (range 131–1230, faster for the peri-operative cases than those that presented to the ED). One post-IAT patient developed a TIA consisting of fluctuating left facial and upper extremity weakness, ataxia, and dysarthria. Another patient developed profuse epistaxis after IAT administration, although it was noteworthy that he was day 1 post-sinus surgery. The median length of stay was 2.8 days, with one patient admitted for 7 days to undergo right carotid artery stenting. All patients were discharged home.

Visual outcomes after thrombolysis

Visual outcomes were analysed for all patients that had at least 3 months follow-up VA available (n = 15) and compared with a cohort of 15 age, gender and baseline VA matched CRAO patients who were treated with conservative approaches obtained from the REP database.

In the IVT cohort, follow-up VA data was available in 66.6% (n = 8) of patients (Figure 1). The previously described CRAO mimic was excluded from this analysis. Mean logMAR VA at presentation was 2.4 ± 0.4 in both the IVT (n = 12) and control groups (n = 12). There was no statistically significant difference in mean logMAR VA at follow-up between the IVT and control groups (2.0 ± 0.9 versus 1.9 ± 0.6, p = .42). Two IVT and three control patients improved by at least two Snellen lines and one patient in each group improved by more than three Snellen lines in the follow-up period. The mean logMAR difference between the initial presentation and follow-up was comparable in the IVT and control groups (0.3 versus 0.4, p = .4).

Figure 1.

Visual outcomes for central retinal artery occlusion patients who received intravenous thrombolysis versus conservative management.

logMAR = logarithm of the minimum angle of resolution; VA = visual acuity.

All IAT patients had follow-up VA data (Figure 2). Baseline mean logMAR VA at admission was 2.6 ± 0.4 in both the IAT and control groups (n = 7 in each). Mean logMAR VA at follow-up was similar between the IAT and control groups (1.9 ± 0.7 versus 2.1 ± 0.8, p = .6). Three IAT patients and two control patients improved by at least two Snellen lines, and two IAT patients and one control patient improved by more than 3 Snellen lines in the follow-up period.

Figure 2.

Visual outcomes for central retinal artery occlusion patients who received intra-arterial therapy versus conservative management.

logMAR = logarithm of the minimum angle of resolution; VA = visual acuity.

Fifty per cent of patients in the IVT group and 71.4% in the IAT group received both thrombolysis and other conservative therapies (Table 2), hence only 8 CRAO patients (42% of total, 6 IVT and 2 IAT) received exclusively thrombolytic therapy. Their follow-up visual outcomes were similar to those of the matched cohort (p > .05 in both comparisons) (Figure 3). No complications were noted in the thrombolysis only group and the mean length of stay was 2.3 ± 1.6 days.

Table 2.

Conservative therapies received by the two matched cohorts of central retinal artery occlusion patients.

Treatment type	IVT group, number, %	IVT control group, number, %	IAT group, number, %	IAT control group, number, %
Thrombolysis	12 (100%)	0 (0%)	7 (100%)	0 (0%)
Verapamil	0 (0%)	0 (0%)	3 (42.86%)	0 (0%)
Ocular massage	4 (33.33%)	3 (37.5%)	2 (28.57%)	2 (28.57%)
HBOT	0 (0%)	2 (25%)	0 (0%)	3 (42.86%)
IOP – lowering drops	4 (33.33%)	1 (12.5%)	3 (42.86%)	1 (14.29%)
Acetazolamide/mannitol	0 (0%)	1 (12.5%)	2 (28.57%)	0 (0%)
Corticosteroids	2 (16.67%)	1(12.5%)	1 (14.29%)	0 (0%)
Paracentesis	0 (0%)	2 (25%)	1 (14.29%)	4 (57.14%)
Pentoxifylline	0 (0%)	0 (0%)	1 (14.29%)	0 (0%)

CRAO = central retinal artery occlusion; HBOT = hyperbaric oxygen therapy; IAT = intra-arterial thrombolysis; IOP = intraocular pressure; IVT = intravenous thrombolysis.

Figure 3.

Visual outcomes for central retinal artery occlusion patients who received thrombolysis alone versus conservative management.

logMAR = logarithm of the minimum angle of resolution; VA = visual acuity.

Discussion

Similar to cerebral AIS, any potential therapeutic intervention for acute retinal ischaemia requires early action. Until a sufficiently powered randomised prospective clinical trial studying early thrombolysis is completed, clinical equipoise in the utility of IVT or IAT for CRAO remains. The present study describes the spectrum of safety and efficacy of thrombolysis, both intravenous and intra-arterial in the hyperacute treatment of CRAO, at our academic multi-centre institution. Despite a sizeable number of patients presenting with acute CRAO, a minority (3.5%) received thrombolysis in the past two decades. Other retrospective studies have reported similar small proportions.²⁴ One observational study of 181 CRAO patients presenting to a tertiary care institution found only 1.6% (n = 3) received thrombolysis (IVT in all). An additional 4.9% (n = 9) of them presented within the thrombolytic window but ultimately were not deemed candidates for this time-sensitive therapy due to delays in diagnosis.²⁵ This highlights the current gap in care and emphasises the need for early recognition and action, so as to maximise positive visual outcomes. Despite tPA approval in 1997, all our CRAO patients received IVT in the past 6 years, reflecting an increased trend towards IVT utilisation in the past decade. However, the limited CRAO presentations in the thombolysis eligible window hinders our overall ability to study this therapy. The FAST stroke awareness mnemonic was expanded to BEFAST (Balance, Eye, Face, Arm, Speech, Time), to indicate that acute visual loss could be caused by an acute stroke, that requires timely specialised evaluation.²⁶ Educational campaigns should continue to target the public and eye care providers to emphasise the need for urgent evaluation of CRAO in specialised stroke centres for consideration of acute reperfusion therapies.

No statistically significant differences were noted in follow-up VA of our thombolysed patients when compared with a cohort of subjects with similar ages, genders and baseline VA treated with non-thrombolytic therapies. Only three patients in the thrombolysis group (1/8 IVT and 2/7 IAT) and two patients in the control group improved by over three Snellen lines in this follow-up period. These results are similar to a recent retrospective observational study comparing IVT (n = 16) to conservative therapies (n = 21), which showed that mean VA at hospital discharge was comparable (logMAR VA 2.1 in IVT group versus 2.3 in conservative therapies group, p = .3).⁵ However, a categorical analysis in that study did reveal that three patients from the IVT group achieved at least reading ability, whereas no patients from the conservative therapies cohort achieved this degree of functionality. Notably, there was no long-term follow-up in this study, and visual outcomes were measured at the time of the hospital discharge, unlike our study that reported the VA at least 3 months after thrombolysis. Conversely, aggregate analyses of multiple studies suggest a greater benefit of thrombolysis. A meta-analysis comparing visual outcomes in patients treated with IVT (n = 147) to those receiving minimal or no treatment (n = 396) found a three-fold greater rate of visual recovery in patients treated with IVT within 4.5 hours of symptom onset (n = 34, of whom 50% achieved visual recovery), with a number needed to treat of four (95% confidence intervals [CI], 2.6–6.6).⁹ Approximately, 17.7% (95% CI, 13.9–21.4%) of those who received minimal/no treatment experienced spontaneous visual recovery. Interestingly, those treated with conservative interventions (ocular massage, anterior chamber paracentesis, haemodilution) were noted to have a lower visual recovery rate of 7.4% (95% CI, 3.7–11.1%), supporting that those therapies may be futile.^9,14 An updated meta-analysis was performed in 2020, identifying an additional 67 patients treated with IVT within 4.5 hours of symptom onset, and noted visual recovery in 37.3% of patients, which was significantly greater than the 17.7% without treatment who experienced spontaneous visual recovery (p = .0005).²⁰

Large scale analyses of visual outcomes in IAT for CRAO are more sparse. A systematic review found 93% of CRAO patients treated with IAT (n = 147) experienced improvement in VA (at least 20/20 in 13%, at least 20/40 in 25%, and at least 20/200 in 41%).²⁷ A 2022 meta-analysis¹⁵ including 8 studies enrolling 269 CRAO patients who received IAT and had baseline and final VA reported, showed that the pooled rate of VA improvement was higher in CRAO patients who received IAT than those without IAT (odds ratio [OR] 4.6, 95% CI [2.1, 10.1], p = .0002; I2 = 0%). A greater OR was reported in IAT within 6 hours from onset to procedure. The benefit remained consistent when VA improvement was defined as ≥3 lines on the Snellen chart and was even greater when CRAO was incomplete. In 103 patients who received IAT within 6 hours, there were no symptomatic intracranial haemorrhages; five TIAs and one AIS were reported. In a retrospective consecutive interventional case series including 15 patients who received IAT within 5.5 and 12 hours post-onset, 53% improved three Snellen lines or more, and 27% improved from CF to >20/80 at 3 weeks.²⁸ Despite positive treatment effects, the potential for selection and publication bias in the above series and meta-analyses must be considered. Additionally, the variability in follow-up in various reports (ranging from days to months) poses limitations regarding the applicability of these results, as CRAO patients can experience spontaneous visual recovery within 1 week of onset as a part of the natural history of the disease.²⁹ Thus, consistent longer-term follow-up visual outcome determination is paramount in accurately assessing the clinical benefit of thrombolysis in this patient population.

The management of acute CRAO in our multi-site academic stroke centre highlights heterogeneity and variability of management, as most patients concomitantly received thrombolysis and a variety of other conservative treatments. More patients in the IAT cohort (71.4%) received additional therapies compared to those that received IVT (50%). The visual outcomes were not significantly different in the small thrombolysis only group (n = 8) compared to the matched group that received conservative therapies.

In regards to safety, the majority of patients who received thrombolysis were ultimately discharged home. The exception was a single patient in the IVT group who developed severe sICH post-alteplase and was discharged to a hospice. It is noteworthy that this patient was an elderly male with a pre-existing diagnosis of dementia and bilateral severe carotid stenosis, though had no identified clinical, laboratory, or radiographic contraindications to thrombolysis, that was administered within 220 minutes of severe monocular visual loss (NLP). In a randomised controlled trial of eight CRAO patients treated with IVT, one patient (12.5%) with underlying cerebral amyloid angiopathy suffered sICH.³⁰ It is necessary to recall that thrombolysis itself poses risk of sICH in acute stroke patients.³¹ Thus, sICH is a rare, but a noted complication of thrombolysis in CRAO. In a 2020 meta-analysis¹⁴ of acute CRAO patients treated with thrombolysis (60 IVT, 134 IAT), four patients (1.9%, 2 IVT, 2 IAT) experienced sICH. In a different retrospective analysis of 30 CRAO patients treated with IVT, one (3.3%) experienced sICH, although this patient also received intravenous heparin after alteplase administration.¹⁷ sICH has also been reported in IAT. In one systematic review of 158 patients, one experienced sICH.²⁷ In a randomised control trial of 42 patients receiving IAT, two (4.76%) developed sICH resulting in hemiparesis that subsequently improved during the study period.³² Another patient in our IAT cohort had a TIA, an ischaemic complication that has been previously reported; of note the CRAO aetiology was embolic in this case. In a retrospective case series of 46 CRAO patients treated with IAT, two (4.35%) developed transient hemiparesis.³³ Aside from epistaxis in a patient who had undergone recent sinus surgery, no other complications of IAT were noted in our cohort. Other complications of thrombolysis noted in the literature include hypertensive crisis, vitreous haemorrhage, haemorrhagic complications of diabetic retinopathy, haematuria, hemilingual oedema, and angioedema.^14,30,34 Patients treated with IAT have also been reported to have carotid vasospasm.³²

One patient who received IVT after the initial neurology and before ophthalmology evaluation was deemed to be a CRAO mimic, who had acute recognition of vision loss from cataract. No complications were encountered in this patient. This highlights the importance of developing institutional protocols that allow rapid ophthalmological examination of possible NA-CRAO patients, either in-person or remotely via non-mydriatic fundoscopic cameras, remote or centralised ophthalmologic fundoscopic image interpretation, and telemedicine.^2,35,36

There are limitations to our study, including its retrospective nature and limited follow-up in some patients. The sample size was small in each treatment group and most patients that received thrombolysis also received additional conservative therapies. Hence, our study was insufficiently powered to detect differences in visual outcomes between thrombolysis alone and other heteregenous conservative therapies. We could not determine if etiological subgroups have dissimilar outcomes after thrombolysis alone, given small numbers in each subgroup. Additionally, only one patient received tenecteplase, and despite this agent’s recent translation in some stroke centres as the intravenous thrombolytic of choice,^37–40 its role in CRAO remains unknown.

Conclusion

Our study showed limited utilisation of thrombolysis in patients with CRAO and the routine use of various conservative therapies in addition to thrombolysis. In the absence of level 1 evidence data, we recommend considering a shared-decision practice, with in-depth counselling of the patients and families and careful weighing of the expected benefits and small potential risks of early thrombolysis in select NA-CRAO cases presenting with disabling visual loss. Given the similar outcomes in our cohort receiving thrombolysis and not receiving thrombolysis, prospective studies comparing early thrombolysis and placebo (or anti-thrombotic agent) are warranted to guide hyperacute CRAO management.

Funding Statement

This work was not supported by any funding.

Disclosure statement

No potential conflict of interest was reported by the author(s).