- Email: [email protected]
Prevalence of Retinal Emboli and Acute Retinal Artery Occlusion in Acute Ischemic Stroke
ARTICLE IN PRESS
Prevalence of Retinal Emboli and Acute Retinal Artery Occlusion in Acute Ischemic Stroke Robert A. Egan, MD,* and Helmi L. Lutsep, MD†
Objective: In population-based studies asymptomatic retinal emboli occur in .32%2.9% of people. Retinal artery occlusion (RAO) may occur concurrently with cerebral stroke but the frequency is unknown. No study has examined how commonly retinal emboli occur in the acute stroke population. We aimed to assess the prevalence of retinal emboli and RAO at the time of carotid territory ischemic stroke. Methods: Patients were enrolled prospectively after onset of symptoms consistent with the diagnosis of carotid territory ischemic stroke. Every participant underwent pharmacologic dilation of both pupils and bedside funduscopic examination. Emboli were classified as cholesterol, calcific, platelet/fibrin, or other and categorized by the side of occurrence. Stroke was classified as atheroembolic, cardioembolic, embolic stroke of undetermined source, lacunar, or other. Acute RAO was diagnosed by direct visualization of ischemic retinal whitening. Results: Sixty-five patients were enrolled with a mean age of 59.2 years; 23 were female (35.4%). Eleven of 65 subjects (16.9%) had retinal emboli visible on funduscopy; all were cholesterol emboli except a single platelet/fibrin embolus in a patient with atheroembolic source. Six patients (9%) had acute RAO and no RAO was seen in the lacunar or undetermined source subgroups. Conclusions: Retinal emboli occurred more than 10 times more frequently in the acute stroke patient than in large population-based studies. RAOs also occurred concurrently with ischemic stroke. Although emboli were seen in patients with atheroembolic and cardioembolic sources, all patients with carotid disease had emboli in the ipsilateral eye. Future studies are required to determine if the presence of retinal emboli or RAO may help elucidate an etiology in patients suffering from embolic stroke of undetermined source. Key Words: Retinal artery occlusion—retinal emboli—ischemic stroke— Hollenhorst plaque © 2019 Elsevier Inc. All rights reserved.
Introduction Retinal emboli are particulate intravascular matter that travel to the eye through the central retinal artery and may cause retinal artery occlusion (RAO). Typically they consist of cholesterol (Hollenhorst plaques), platelet-fibrin debris (Fisher plaques), or calcium. The prevalence in population studies has been reported between 0.32% and 2.9%.1-8 However, most observational population based
studies make no distinction between the different types of emboli. It is also unknown what the prevalence of RAO is in the acute ischemic stroke population although there is some data regarding the reverse: the prevalence of stroke in a population of RAO patients has been reported at 24%.9 The incidence of RAO in the general population has been estimated at 1 to 2 per 100,000 people per year.10 Also, a Taiwanese study found that nearly 20% of RAO patients suffered stroke within 3 years of their incident
From the *MultiCare Rockwood Clinic, Department of Neurology and Ophthalmology, Spokane, Washington; and †Oregon Stroke Center, Department of Neurology, Oregon Health and Science University, Portland, Oregon. Received August 5, 2019; revision received September 19, 2019; accepted September 22, 2019. Funding: None. Address correspondence to Dr. Robert A. Egan, MD, MultiCare Rockwood Clinic, 910 W 5th Ave, Suite 1000, Spokane, WA 99204. E-mail: [email protected]. 1052-3057/$ - see front matter © 2019 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jstrokecerebrovasdis.2019.104446
Journal of Stroke and Cerebrovascular Diseases, Vol. &&, No. && (&&), 2019: 104446
1
ARTICLE IN PRESS R.A. EGAN AND H.L. LUTSEP
2
ocular event.11 A Korean study found that 8% of patients who suffered a central RAO had an ischemic stroke within 1 year prior to or after their incident ocular event.12 Finally, a retrospective review of 300 subjects revealed that the risk of stroke following central RAO was much higher if the embolic source was known as opposed to being undertemined.13 The 3 common etiologies of embolic stroke include atherosclerotic, cardioembolic, and cryptogenic or what is now termed embolic stroke of undetermined source (ESUS). It is currently unknown what the exact prevalence of retinal emboli in these types of stroke is but they have been reported more commonly in stroke from atherosclerotic etiology.3,4 More importantly, there is no data on prevalence of emboli in the acute ischemic stroke patient overall. Data about the prevalence of retinal emboli in stroke subtypes may help practitioners determine a more definitive stroke etiology especially when confronted with those with ESUS. We conducted a pilot study of consecutive acute stroke patients to determine if prevalence of retinal emboli mirrors that of population studies of asymptomatic patients using bedside direct ophthalmoscopy. We also strove to determine the frequency of acute RAO in the ischemic stroke population.
Materials and Methods This research was reviewed and approved by the Oregon Health & Science University Institutional Review Board. Subjects were enrolled prospectively after onset of symptoms consistent with the diagnosis of carotid territory ischemic stroke. Lacunar strokes were included if they were not in the posterior circulation. Every participant underwent pharmacologic dilation of both pupils and funduscopy was performed with a direct ophthalmoscope by the same examiner (R.A.E.). Emboli were classified as cholesterol if noted to be orange and refractile and located at retinal arteriolar bifurcations, calcific if white and located on or near
the optic disc, and platelet/fibrin if grey to yellow in appearance and more longitudinally extensive. Emboli were listed as other if they did not fit these criteria. They were also categorized by the side of occurrence. RAO was diagnosed by direct visualization of ischemic retinal whitening on funduscopy. Stroke subtypes were identified as atheroembolic, cardioembolic, ESUS, lacunar, or other. No patient had extended post hospital cardiac monitoring and subjects were considered to have exited the study period at discharge from the hospital. Etiologies such as dissection, hypercoagulable state, and Moya Moya syndrome were placed in the Other category. Risk factors including hypertension, diabetes mellitus, coronary artery disease, prior ischemic stroke, smoking, and hypercholesterolemia were recorded.
Results Sixty-five consecutive patients were enrolled with a mean age of 59.2 years (19-94); 23 were female (35.4%). The mean National Institutes of Health Stroke Scale (NIHSS) was 9.9 (0-27) and patients were examined 2.8 days after their stroke (R = 0-21 days). Nine strokes were lacunar with 1 lacunar stroke being related to Fabry’s disease; the remaining 56 strokes were embolic. Forty-two subjects (65%) identified as past smokers and 24 (37%) were currently smoking. Eleven of 65 subjects (16.9%) had retinal emboli visible on funduscopy; all were cholesterol emboli except a single platelet/fibrin embolus in a patient with atherosclerotic disease. No patient had calcific or talc emboli, pseudoemboli, or other types of emboli. Four of the 20 patients with atheroembolic source had emboli and all were in the eye ipsilateral to the implicated artery. Those patients with cardioembolic disease or ESUS had emboli on either side. No patient with a lacunar stroke or “Other” stroke had a retinal embolus. Six patients (9%) had acute RAO and no RAO was seen in the lacunar or ESUS groups. All RAO were branch RAO and not central RAO. These results are summarized in Table 1.
Table 1. Characteristics of subjects regarding etiology
Number Female Mean age Mean NIHSS # with emboli >1 embolus Emboli, ipsi RAO
Athero
Cardio
ESUS
Lacunar
Other
20 6/20 (30%) 60 8.75 4 2/4 4/4 3/20
15 4/15 (27%) 62 11.60 3 3/3 2/3 2/15
12 5/12 (42%) 59 11.92 4 2/4 2/4 0/12
9 2/9 (22%) 61 5.33 0 0 0/0 0/9
9 5/9 (55%) 49 11.56 0 0 0/0 1/9
Abbreviations: ASO, atheroembolic; Cardio, cardioembolic; ESUS, embolic stroke of undetermined source; ipsi, ipsilateral; RAO, retinal artery occlusion. All emboli were cholesterol except for a single embolus that was platelet/fibrin in the ASO category. One patient with stroke of cardioembolic type had bilateral emboli. The single patient of Other etiology with retinal artery occlusion had the antiphospholipid antibody syndrome.
ARTICLE IN PRESS RETINAL EMBOLI AND RAO IN STROKE
Discussion We showed that retinal emboli are present in acute stroke patients at a much higher frequency than the asymptomatic population that has been reported in the past. In our atheroembolic group, emboli were always found on the ipsilateral side to the carotid stenosis. We also demonstrated that acute RAO occurs at the time of presentation of acute stroke and more frequently than in population based studies. RAO and retinal emboli did not occur in lacunar stroke. Also, RAO occurred in 2 of our acute stroke patients with atrial fibrillation, a stroke etiology that has been downplayed in the past.15 It is technically difficult to study the incidence of various retinal emboli in acute stroke as patients are typically not presenting to their ophthalmologist prior to their cerebrovascular event and therefore most data can really only relate to the prevalence of such emboli. We therefore cannot state with any certainty that the retinal emboli present on our subjects’ examinations occurred at any time near their incident acute stroke or whether they occurred years prior. However, we did show that acute RAO does occur near the incident ischemic stroke since we based our detection method on visualization of retinal ischemic whitening, a sign of acute retinal infarction that disappears shortly after onset, rather than using visual fields or multifocal electroretinograms which may reveal abnormalities from retinal infarction that may have occurred years prior. One can argue that another limitation of our study was that subjects were not examined by indirect ophthalmoscopy or fundus photography. We used direct ophthalmoscopy and some of our subjects were difficult to examine as they were not fully cooperative given their acute neurologic syndrome. It is conceivable that we missed some emboli and RAO due to this manner of examination. The reason for the decision to use direct ophthalmoscopy was to attempt to create a research milieu that most closely resembled the real physiological world where practitioners would be evaluating patients at the bedside without portable fundus cameras. There is a push now to equip emergency departments with fundus cameras to be used in patients with focal neurologic deficits but this is only present in a few centers at this time.14 Future studies should investigate the rate and types of retinal emboli in stroke of different etiologies including carotid atherosclerotic disease, atrial fibrillation, and ESUS by examining larger populations and should investigate whether side predominance is related only to the atheroembolic group; in our subjects with embolic stroke that was not atheroembolic, retinal emboli were found ipsilateral or contralateral to the ischemic stroke. This initial study also confirms that RAO occurs simultaneously with acute stroke not infrequently. We saw RAO in both the atheroembolic and cardioembolic groups and larger numbers are required to determine any other specific differences between these populations. Future research
3
should also be aimed at determining if the presence of RAO at onset of acute ischemic stroke indicates that the etiology is atheroembolic or cardioembolic rather than from ESUS.
Conflict of interest The Authors have no conflict of interest.
References 1. Cheung N, Teo K, Zhao W, et al. Prevalence and associations of retinal emboli with ethnicity, stroke, and renal disease in a multiethnic Asian population: The Singapore Epidemiology of Eye Disease Study. JAMA Ophthalmol 2017;135:1023-1028. 2. Klein R, Klein BEK, Jensen SC, et al. Retinal emboli and stroke: The Beaver Dam Eye Study. Arch Ophthalmol 1999;117:1063-1068. 3. Ahmed R, Khetpal V, Merin LM, et al. Case series: retrospective review of incidental retinal emboli found on diabetic retinopathy screening: is there a benefit to referral for work-up and possible management? Clin Diabetes 2008;26:179-182. 4. Hadley G, Earnshaw JJ, Stratton I, et al. A potential pathway for managing diabetic patients with arterial emboli detected by retinal screening. Eur J Vasc Endovasc Surg 2011;42:153-157. 5. Cheung N, Lim L, Wang JJ, et al. Prevalence and risk factors of retinal arteriolar emboli: The Singapore Malay Eye Study. Am J Ophthalmol 2008;146:620-624. 6. Hoki SL, Varma R, Lai MY, et al. Prevalence and associations of asymptomatic retinal emboli in latinos: The Los Angeles Latino Eye Study (LALES). Am J Ophthalmol 2008;145:143-148. 7. Mitchell P, Wang JJ, Li W, et al. Prevalence of asymptomatic retinal emboli in an Australian urban community. Stroke 1997;28:63-66. 8. Cugati S, Wang JJ, Rochtchina E, et al. Ten-Year incidence of retinal emboli in an older population. Stroke 2006;37:908-910. 9. Lee J, Kim SW, Lee SC, et al. Co-occurrence of acute retinal artery occlusion and acute ischemic stroke: diffusionweighted magnetic resonance imaging study. Am J Ophthalmol 2014;157:1231-1238. 10. Leavitt JA, Larson TA, Hodge DO, et al. The incidence of central retinal artery occlusion in olmstead County, Minnesota. Am J Ophthalmol 2011;152:820-823. 11. Chang YS, Jan RL, Weng SF, et al. Retinal artery occlusion and the 3-year risk of stroke in Taiwan: a nationwide population-based study. Am J Ophthalmol 2012; 154:645-652. 12. Park SJ, Cho NK, Yang BR, et al. Risk and risk periods for stroke and acute myocardial infarction in patients with central retinal artery occlusion. Ophthalmology 2015;122:2336-2343. 13. Chodnicki KD, Pulido JS, Hodge DO, et al. Stroke risk before and after central retinal artery occlusion in a US Cohort. Mayo Clin Proc 2019;94:236-241. 14. Vuong LN, Thulasi P, Biousse V, et al. Ocular fundus photography of patients with focal neurologic deficits in an emergency department. Neurology 2015;85:256-262. 15. Zarkali A, Cheng SF, Dados A, et al. Atrial fibrillation: An Underestimated Cause of Ischemic Monocular Visual Loss? J Stroke Cerebrovasc Dis 2019;28:1495-1499.
Prevalence of Retinal Emboli and Acute Retinal Artery Occlusion in Acute Ischemic Stroke Robert A. Egan, MD,* and Helmi L. Lutsep, MD†
Objective: In population-based studies asymptomatic retinal emboli occur in .32%2.9% of people. Retinal artery occlusion (RAO) may occur concurrently with cerebral stroke but the frequency is unknown. No study has examined how commonly retinal emboli occur in the acute stroke population. We aimed to assess the prevalence of retinal emboli and RAO at the time of carotid territory ischemic stroke. Methods: Patients were enrolled prospectively after onset of symptoms consistent with the diagnosis of carotid territory ischemic stroke. Every participant underwent pharmacologic dilation of both pupils and bedside funduscopic examination. Emboli were classified as cholesterol, calcific, platelet/fibrin, or other and categorized by the side of occurrence. Stroke was classified as atheroembolic, cardioembolic, embolic stroke of undetermined source, lacunar, or other. Acute RAO was diagnosed by direct visualization of ischemic retinal whitening. Results: Sixty-five patients were enrolled with a mean age of 59.2 years; 23 were female (35.4%). Eleven of 65 subjects (16.9%) had retinal emboli visible on funduscopy; all were cholesterol emboli except a single platelet/fibrin embolus in a patient with atheroembolic source. Six patients (9%) had acute RAO and no RAO was seen in the lacunar or undetermined source subgroups. Conclusions: Retinal emboli occurred more than 10 times more frequently in the acute stroke patient than in large population-based studies. RAOs also occurred concurrently with ischemic stroke. Although emboli were seen in patients with atheroembolic and cardioembolic sources, all patients with carotid disease had emboli in the ipsilateral eye. Future studies are required to determine if the presence of retinal emboli or RAO may help elucidate an etiology in patients suffering from embolic stroke of undetermined source. Key Words: Retinal artery occlusion—retinal emboli—ischemic stroke— Hollenhorst plaque © 2019 Elsevier Inc. All rights reserved.
Introduction Retinal emboli are particulate intravascular matter that travel to the eye through the central retinal artery and may cause retinal artery occlusion (RAO). Typically they consist of cholesterol (Hollenhorst plaques), platelet-fibrin debris (Fisher plaques), or calcium. The prevalence in population studies has been reported between 0.32% and 2.9%.1-8 However, most observational population based
studies make no distinction between the different types of emboli. It is also unknown what the prevalence of RAO is in the acute ischemic stroke population although there is some data regarding the reverse: the prevalence of stroke in a population of RAO patients has been reported at 24%.9 The incidence of RAO in the general population has been estimated at 1 to 2 per 100,000 people per year.10 Also, a Taiwanese study found that nearly 20% of RAO patients suffered stroke within 3 years of their incident
From the *MultiCare Rockwood Clinic, Department of Neurology and Ophthalmology, Spokane, Washington; and †Oregon Stroke Center, Department of Neurology, Oregon Health and Science University, Portland, Oregon. Received August 5, 2019; revision received September 19, 2019; accepted September 22, 2019. Funding: None. Address correspondence to Dr. Robert A. Egan, MD, MultiCare Rockwood Clinic, 910 W 5th Ave, Suite 1000, Spokane, WA 99204. E-mail: [email protected]. 1052-3057/$ - see front matter © 2019 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.jstrokecerebrovasdis.2019.104446
Journal of Stroke and Cerebrovascular Diseases, Vol. &&, No. && (&&), 2019: 104446
1
ARTICLE IN PRESS R.A. EGAN AND H.L. LUTSEP
2
ocular event.11 A Korean study found that 8% of patients who suffered a central RAO had an ischemic stroke within 1 year prior to or after their incident ocular event.12 Finally, a retrospective review of 300 subjects revealed that the risk of stroke following central RAO was much higher if the embolic source was known as opposed to being undertemined.13 The 3 common etiologies of embolic stroke include atherosclerotic, cardioembolic, and cryptogenic or what is now termed embolic stroke of undetermined source (ESUS). It is currently unknown what the exact prevalence of retinal emboli in these types of stroke is but they have been reported more commonly in stroke from atherosclerotic etiology.3,4 More importantly, there is no data on prevalence of emboli in the acute ischemic stroke patient overall. Data about the prevalence of retinal emboli in stroke subtypes may help practitioners determine a more definitive stroke etiology especially when confronted with those with ESUS. We conducted a pilot study of consecutive acute stroke patients to determine if prevalence of retinal emboli mirrors that of population studies of asymptomatic patients using bedside direct ophthalmoscopy. We also strove to determine the frequency of acute RAO in the ischemic stroke population.
Materials and Methods This research was reviewed and approved by the Oregon Health & Science University Institutional Review Board. Subjects were enrolled prospectively after onset of symptoms consistent with the diagnosis of carotid territory ischemic stroke. Lacunar strokes were included if they were not in the posterior circulation. Every participant underwent pharmacologic dilation of both pupils and funduscopy was performed with a direct ophthalmoscope by the same examiner (R.A.E.). Emboli were classified as cholesterol if noted to be orange and refractile and located at retinal arteriolar bifurcations, calcific if white and located on or near
the optic disc, and platelet/fibrin if grey to yellow in appearance and more longitudinally extensive. Emboli were listed as other if they did not fit these criteria. They were also categorized by the side of occurrence. RAO was diagnosed by direct visualization of ischemic retinal whitening on funduscopy. Stroke subtypes were identified as atheroembolic, cardioembolic, ESUS, lacunar, or other. No patient had extended post hospital cardiac monitoring and subjects were considered to have exited the study period at discharge from the hospital. Etiologies such as dissection, hypercoagulable state, and Moya Moya syndrome were placed in the Other category. Risk factors including hypertension, diabetes mellitus, coronary artery disease, prior ischemic stroke, smoking, and hypercholesterolemia were recorded.
Results Sixty-five consecutive patients were enrolled with a mean age of 59.2 years (19-94); 23 were female (35.4%). The mean National Institutes of Health Stroke Scale (NIHSS) was 9.9 (0-27) and patients were examined 2.8 days after their stroke (R = 0-21 days). Nine strokes were lacunar with 1 lacunar stroke being related to Fabry’s disease; the remaining 56 strokes were embolic. Forty-two subjects (65%) identified as past smokers and 24 (37%) were currently smoking. Eleven of 65 subjects (16.9%) had retinal emboli visible on funduscopy; all were cholesterol emboli except a single platelet/fibrin embolus in a patient with atherosclerotic disease. No patient had calcific or talc emboli, pseudoemboli, or other types of emboli. Four of the 20 patients with atheroembolic source had emboli and all were in the eye ipsilateral to the implicated artery. Those patients with cardioembolic disease or ESUS had emboli on either side. No patient with a lacunar stroke or “Other” stroke had a retinal embolus. Six patients (9%) had acute RAO and no RAO was seen in the lacunar or ESUS groups. All RAO were branch RAO and not central RAO. These results are summarized in Table 1.
Table 1. Characteristics of subjects regarding etiology
Number Female Mean age Mean NIHSS # with emboli >1 embolus Emboli, ipsi RAO
Athero
Cardio
ESUS
Lacunar
Other
20 6/20 (30%) 60 8.75 4 2/4 4/4 3/20
15 4/15 (27%) 62 11.60 3 3/3 2/3 2/15
12 5/12 (42%) 59 11.92 4 2/4 2/4 0/12
9 2/9 (22%) 61 5.33 0 0 0/0 0/9
9 5/9 (55%) 49 11.56 0 0 0/0 1/9
Abbreviations: ASO, atheroembolic; Cardio, cardioembolic; ESUS, embolic stroke of undetermined source; ipsi, ipsilateral; RAO, retinal artery occlusion. All emboli were cholesterol except for a single embolus that was platelet/fibrin in the ASO category. One patient with stroke of cardioembolic type had bilateral emboli. The single patient of Other etiology with retinal artery occlusion had the antiphospholipid antibody syndrome.
ARTICLE IN PRESS RETINAL EMBOLI AND RAO IN STROKE
Discussion We showed that retinal emboli are present in acute stroke patients at a much higher frequency than the asymptomatic population that has been reported in the past. In our atheroembolic group, emboli were always found on the ipsilateral side to the carotid stenosis. We also demonstrated that acute RAO occurs at the time of presentation of acute stroke and more frequently than in population based studies. RAO and retinal emboli did not occur in lacunar stroke. Also, RAO occurred in 2 of our acute stroke patients with atrial fibrillation, a stroke etiology that has been downplayed in the past.15 It is technically difficult to study the incidence of various retinal emboli in acute stroke as patients are typically not presenting to their ophthalmologist prior to their cerebrovascular event and therefore most data can really only relate to the prevalence of such emboli. We therefore cannot state with any certainty that the retinal emboli present on our subjects’ examinations occurred at any time near their incident acute stroke or whether they occurred years prior. However, we did show that acute RAO does occur near the incident ischemic stroke since we based our detection method on visualization of retinal ischemic whitening, a sign of acute retinal infarction that disappears shortly after onset, rather than using visual fields or multifocal electroretinograms which may reveal abnormalities from retinal infarction that may have occurred years prior. One can argue that another limitation of our study was that subjects were not examined by indirect ophthalmoscopy or fundus photography. We used direct ophthalmoscopy and some of our subjects were difficult to examine as they were not fully cooperative given their acute neurologic syndrome. It is conceivable that we missed some emboli and RAO due to this manner of examination. The reason for the decision to use direct ophthalmoscopy was to attempt to create a research milieu that most closely resembled the real physiological world where practitioners would be evaluating patients at the bedside without portable fundus cameras. There is a push now to equip emergency departments with fundus cameras to be used in patients with focal neurologic deficits but this is only present in a few centers at this time.14 Future studies should investigate the rate and types of retinal emboli in stroke of different etiologies including carotid atherosclerotic disease, atrial fibrillation, and ESUS by examining larger populations and should investigate whether side predominance is related only to the atheroembolic group; in our subjects with embolic stroke that was not atheroembolic, retinal emboli were found ipsilateral or contralateral to the ischemic stroke. This initial study also confirms that RAO occurs simultaneously with acute stroke not infrequently. We saw RAO in both the atheroembolic and cardioembolic groups and larger numbers are required to determine any other specific differences between these populations. Future research
3
should also be aimed at determining if the presence of RAO at onset of acute ischemic stroke indicates that the etiology is atheroembolic or cardioembolic rather than from ESUS.
Conflict of interest The Authors have no conflict of interest.
References 1. Cheung N, Teo K, Zhao W, et al. Prevalence and associations of retinal emboli with ethnicity, stroke, and renal disease in a multiethnic Asian population: The Singapore Epidemiology of Eye Disease Study. JAMA Ophthalmol 2017;135:1023-1028. 2. Klein R, Klein BEK, Jensen SC, et al. Retinal emboli and stroke: The Beaver Dam Eye Study. Arch Ophthalmol 1999;117:1063-1068. 3. Ahmed R, Khetpal V, Merin LM, et al. Case series: retrospective review of incidental retinal emboli found on diabetic retinopathy screening: is there a benefit to referral for work-up and possible management? Clin Diabetes 2008;26:179-182. 4. Hadley G, Earnshaw JJ, Stratton I, et al. A potential pathway for managing diabetic patients with arterial emboli detected by retinal screening. Eur J Vasc Endovasc Surg 2011;42:153-157. 5. Cheung N, Lim L, Wang JJ, et al. Prevalence and risk factors of retinal arteriolar emboli: The Singapore Malay Eye Study. Am J Ophthalmol 2008;146:620-624. 6. Hoki SL, Varma R, Lai MY, et al. Prevalence and associations of asymptomatic retinal emboli in latinos: The Los Angeles Latino Eye Study (LALES). Am J Ophthalmol 2008;145:143-148. 7. Mitchell P, Wang JJ, Li W, et al. Prevalence of asymptomatic retinal emboli in an Australian urban community. Stroke 1997;28:63-66. 8. Cugati S, Wang JJ, Rochtchina E, et al. Ten-Year incidence of retinal emboli in an older population. Stroke 2006;37:908-910. 9. Lee J, Kim SW, Lee SC, et al. Co-occurrence of acute retinal artery occlusion and acute ischemic stroke: diffusionweighted magnetic resonance imaging study. Am J Ophthalmol 2014;157:1231-1238. 10. Leavitt JA, Larson TA, Hodge DO, et al. The incidence of central retinal artery occlusion in olmstead County, Minnesota. Am J Ophthalmol 2011;152:820-823. 11. Chang YS, Jan RL, Weng SF, et al. Retinal artery occlusion and the 3-year risk of stroke in Taiwan: a nationwide population-based study. Am J Ophthalmol 2012; 154:645-652. 12. Park SJ, Cho NK, Yang BR, et al. Risk and risk periods for stroke and acute myocardial infarction in patients with central retinal artery occlusion. Ophthalmology 2015;122:2336-2343. 13. Chodnicki KD, Pulido JS, Hodge DO, et al. Stroke risk before and after central retinal artery occlusion in a US Cohort. Mayo Clin Proc 2019;94:236-241. 14. Vuong LN, Thulasi P, Biousse V, et al. Ocular fundus photography of patients with focal neurologic deficits in an emergency department. Neurology 2015;85:256-262. 15. Zarkali A, Cheng SF, Dados A, et al. Atrial fibrillation: An Underestimated Cause of Ischemic Monocular Visual Loss? J Stroke Cerebrovasc Dis 2019;28:1495-1499.