Influence of Carotid Artery Stenting on the Retina and Choroid

Abstract

Purpose: The effect of carotid artery stenting in patients with unilateral carotid artery stenosis on the retina and choroid was evaluated using swept-source optical coherence tomography angiography (SS-OCTA).

Methods: SS-OCTA examination was conducted before stenting and 4 days and 3 months after stenting. The retinal nerve fiber layer, ganglion cell–inner plexiform layer (GCIPL), inner nuclear layer, superficial vascular complex (SVC), deep vascular complex (DVC), choroidal vascular volume (CVV), and choroidal vascular index were measured. Repeated-measures analysis of variance was performed to assess the impact of carotid artery stenting on optical coherence tomography angiography (OCTA) metrics.

Results: At baseline, 303 eyes from 160 patients (61.82 ± 9.98 years; 85.29% males) were enrolled. SVC and DVC densities and CVV were lower in ipsilateral eyes (stenosed side) compared to contralateral eyes (all P < 0.05). Four days after stenting, a significant increase was seen in SVC density in ipsilateral eyes (P < 0.05) while a significant increase was seen in CVV in ipsilateral eyes and contralateral eyes (both P < 0.05). Three months after stenting (63 patients with 114 eyes), a significant decrease was seen in the GCIPL thickness of ipsilateral and contralateral eyes (all P < 0.001).

Conclusions: Short term after carotid artery stenting, ipsilateral eyes showed a rapid and significant increase in SVC density and CVV.

Translational Relevance: Optical coherence tomography (OCT)/OCTA measurements may have the potential to detect retinal and choroidal changes after stenting. Future research on the long-term effect of stenting on the retina and choroid will be guided by these findings.

Introduction

Carotid artery stenosis (CAS) directly affects the blood supply to the eyes and causes ocular ischemic syndrome.^1,2 Impairment of blood flow in the carotid artery can affect the blood flow in the ocular circulation and cerebral circulation²; thus, CAS is a prominent risk factor for ophthalmic complications such as retinal artery occlusion³ and cerebral disorders such as ischemic stroke.⁴ 

Carotid artery surgeries such as carotid endarterectomy (CEA) and carotid artery stenting are suggested to be effective in preventing stroke in moderate to severe CAS.⁵ Carotid artery stenting is an alternative to CEA in select patients with high-grade asymptomatic or symptomatic CAS,⁶ such as patients with severe stenosis in other major cerebral arteries or stenosis at an intracranial carotid artery. After carotid artery stenting, accumulating reports^7,8 have shown improved blood flow and/or perfusion of the ophthalmic artery and cerebral vessels. 

The retina and choroid are highly vascularized and thus receive a majority of blood supply in the eye. Evaluation of the retina and choroid after carotid surgery (CEA and/or carotid artery stenting) has been of interest recently. Optical coherence tomography (OCT)/OCT angiography (OCTA) is a noninvasive ophthalmic imaging modality that can visualize retinal microvasculature in different layers of the retina and choroid. OCT/OCTA imaging provides detailed information regarding the retinal and choroidal structure and microvasculature and is sensitive to detect subtle changes that occur in the retina and choroid. To date, accumulating reports^9,10 have shown that OCT/OCTA has the potential to detect structural and microvascular changes in the retina and choroid of patients with CAS compared to controls. The surgical benefits of carotid artery stenting and CEA on the retina and choroid are complicated; some studies showed significant improvement in the structure and vasculature of the retina and choroid after surgery while other studies did not observe any change.^1,2,11 The clinical use of the retina and the choroid as surrogate markers of the internal carotid artery changes after CEA or carotid artery stenting is inconclusive given inconsistent findings.^1,2,12 These inconsistencies may be due to various limitations associated with the time interval after surgery, OCT/OCTA tools, software used to quantify the retina and choroid, and the inclusion criteria of enrolled patients. Given these vicissitudes, we conducted a more extensive investigation with larger sample size to explore the impact of carotid stenting on the retina and choroid. 

OCT/OCTA metrics, including retinal structural thicknesses and microvasculature and choroidal parameters (choroidal vascular volume/index), were imaged and measured before and 4 days and 3 months after carotid artery stenting. Imaging was done on both eyes, and measurement were done automatically by the OCT/OCTA tool to elucidate the impact of carotid artery stenting on the retina and choroid. 

Patients and Methods

Study Design and Participants

This study is part of an ongoing cohort study (ChiCTR2300074640) that consecutively recruits patients with carotid atherosclerotic stenosis at the Neurology Department of West China Hospital from April 2021 to December 2022. The rationale and details of the cohort have appeared earlier.¹⁰ Briefly, carotid stenosis/occlusion was screened by computed tomography angiography, magnetic resonance angiography, or ultrasonography and confirmed by digital subtraction angiography (DSA). Here, we focused on unilateral patients with CAS who underwent stenting. 

Patients with multiple severe stenosis in anterior circulation, with other neurologic and vascular diseases (such as Parkinson disease, artery dissection), or nonatherosclerotic stenosis were excluded. Patients with stenting covering the origin of the ophthalmic artery (OA) or distal to OA or with central/branchial retinal artery occlusion were also excluded. 

Procedures were performed following the Declaration of Helsinki. The West China Hospital Research Ethics Committee (2020(922)) approved the study, and written informed consent was acquired from all participants. 

Clinical Features

Gender, age, vascular risk factors (history of hypertension, dyslipidemia, diabetes mellitus, drinking, and smoking), and resting blood pressure were recorded before DSA. Medical history and physical examination of the nervous system were recorded to assess whether a patients with CAS was symptomatic. Symptomatic CAS refers to patients with a history of cerebral infarction, transient ischemic attack, or transient mono-blindness.⁴ 

DSA and Artery Evaluation

DSA was performed to determine the indication for a stenting procedure and assess cerebral artery stenosis. Further, computed tomography perfusion was conducted for judgment of stenting indication in asymptomatic patients, with intracranial CAS or with carotid occlusion. The interarterial DSA images were recorded in both anterior-posterior and lateral projections for bilateral internal carotid artery, vertebral artery, and subclavian artery, and further projections were recorded when necessary to assess intracranial artery stenosis. Poststenting DSA imaging was also performed in the two projections for the operated side to evaluate whether stenosis remained or had distal cerebral embolism. 

The Warfarin-Aspirin Symptomatic Intracranial Disease Study criteria¹³ were used to evaluate stenotic degree for intracranial carotid stenosis, while the North American Symptomatic Carotid Endarterectomy Trial criteria¹⁴ were used for extracranial carotid stenosis. The absence of distal contrast filling was referred to as occlusion. The extracranial part of the carotid artery refers to the common carotid artery and the cervical segment of the internal carotid artery. The eye on the same side as the stenting was defined as the ipsilateral side, while another eye was the contralateral side. 

Retinal Imaging and OCTA Processing

All participants underwent fundus camera photography (VISUCAM 200) and OCT/OCTA imaging to image the optic nerve head (ONH) and fundus and/or retina. Fundus photography and OCT/OCTA imaging was performed 1 to 3 days before and 4 days after stenting. After imaging, an ophthalmologist evaluated all images. A number of our study participants recruited from January 2022 to December 2022 completed further retinal imaging 3 months after stenting. 

Swept-source OCT/OCTA (SS-OCT/OCTA) was used to image the retina and choroid for our study participants. The tool (VG200S, version 2.1.016; SVision Imaging, Henan, China) was equipped with eye-tracking software to eliminate eye motion artifacts. The specifications of the OCT/OCTA tool are well described in our previous reports.^10,15 

Structural OCT imaging of the macular region was performed with 18 radial scans centered on the fovea. Each scan line, generated by 2048 A-scans, was 12 mm long. Automated segmentation of the retinal thickness was performed with the OCT tool; the retinal nerve fiber layer (RNFL), ganglion cell–inner plexiform layer (GCIPL), and inner nuclear layer (INL) of the retina in 6 × 6 mm around the fovea were segmented and imaged. The choroidal vascular volume (CVV) was defined as the volume from the basal border of the retinal pigment epithelium–Bruch membrane complex to the choroidoscleral junction. Choroidal vascular index (CVI) was described as the ratio of the choroidal vascular luminal volume to the total choroidal volume, as shown in Figure 1. Mean thicknesses of the RNFL, GCIPL, INL, CVV, and CVI were automatically measured by the OCT tool. 

OCTA images were obtained with a raster scan protocol that covered 6 × 6 mm around the fovea. The enface angiograms of the superficial vascular complex (SVC) and deep vascular complex (DVC) were generated by automatic segmentation to assess the retinal densities, as shown in Figure 1. The segmentation of the SVC and DVC was set in the inner two-thirds and outer one-third border of the GCIPL, as shown in Figure 1. 

Our exclusion criteria for eyes were as follows: eyes with optic head pallor, retinal hemorrhages, exudate, aneurysms, severe cataracts, and retinoschisis. OCT/OCTA images with retinal pathologies, such as drusen, macular edema, choroidal neovascularization, geographic atrophy, epiretinal membrane, and macular hole, were also excluded. If a participant presented with any of these disorders in one eye, the other eye was used; if both eyes had retinal disorders, the patient was excluded. OCT/OCTA images with a signal quality of less than 7 were also excluded from our study. OCT/OCTA data in our study followed the OSCAR-IB quality criteria¹⁶ and APOSTEL recommendation.¹⁷ 

Statistical Analysis

Demographics and clinical characteristics are described as mean ± standard deviation or median [interquartile range] for continuous variables and frequencies with percentages for categorical variables. When comparing the characteristics between patients with OCT/OCTA examination before and 4 days after stenting and patients with further OCT/OCTA examination at 3 months after stenting, the Fisher exact test, the t-test, or the Kruskal–Wallis test was used as appropriate. The paired t-test was used to compare OCT/OCTA metrics between ipsilateral eyes and contralateral eyes before stenting, 4 days after stenting, or 3 months after stenting. The two eyes in one person at every time point were the pair, considering individual differences in eyes. Repeated-measures analysis of variance (ANOVA) was conducted when comparing OCT/OCTA metrics between eyes before and 4 days after stenting. Repeated-measures ANOVA was also used to compare eyes in the ipsilateral and contralateral subgroup or analysis for comparison between eyes before and 3 months after stenting. Repeated-measures ANOVA was chosen because of the high correlation between OCT/OCTA metrics at different time points. Multiple imputation was conducted to impute missing OCT/OCTA values for the one eye caused by ophthalmic diseases in our enrolled participants. The rationale and details of multiple imputations are shown in the supplementary material. P < 0.05 was considered significant. All analysis and plotting were conducted in R 4.2.3, and mice packages were used for multiple imputations. 

Results

In total, 228 patients with CAS were screened before stenting in our study. Sixty-eight patients were excluded because of nonatherosclerotic stenosis, out of indication of stenting, ophthalmic diseases, or rejection of repeated OCT/OCTA examination, as shown in Figure 2. The final analysis included 160 patients (61.82 ± 9.98 years; 85.29% males) who completed the follow-up 4 days after stenting; 156 ipsilateral eyes and 147 contralateral eyes were included. Of the 160 CAS cohort, 78 patients (48.75%) were asymptomatic and 82 (51.25%) were symptomatic; 139 patients (86.88%) had extracranial stenosis while 21 (13.12%) had intracranial stenosis. The mean stenotic degree was 81.75% in the 160 CAS cohort and 61 patients (38.13%) had severe stenosis in other major cerebral arteries. Sixty-three patients (61.81 ± 10.19 years; 78.84% males) completed further examination at 3 months after stenting; 59 ipsilateral eyes and 55 contralateral eyes of the 63 patients with CAS were included. The Table displays the demographics and clinical information of our study participants. 

Figure 1.

 

Retina imaging evaluation. The segmentation of the SVC and DVC was set in the inner two-thirds and outer one-third border of the GCIPL.

Figure 2.

 

Flowcharts of patient inclusion.

Table.

 

The Clinical Features and Cerebral Artery Information for the Included Patients

 

Supplementary Table S1 shows the comparison of OCT/OCTA metrics between ipsilateral and contralateral eyes. On baseline, ipsilateral eyes showed thinner RNFL thickness and lower SVC density, DVC density, and CVV (P < 0.05) compared to contralateral eyes. No significant differences were shown when OCT/OCTA metrics were compared between ipsilateral and contralateral eyes 4 days after stenting. Three months after stenting, ipsilateral eyes showed thinner GCIPL thickness (P = 0.006) when compared to contralateral eyes. 

Figure 3 shows OCT/OCTA metrics changes for ipsilateral and contralateral eyes after stenting. Four days after stenting, ipsilateral eyes showed increased SVC density (P = 0.008), increased CVV (P < 0.001), and decreased GCIPL thickness (P = 0.018). Increased RNFL thickness (P < 0.001), increased CVV (P = 0.037), and decreased GCIPL thickness (P < 0.001) were seen in contralateral eyes. Three months after stenting, significant GCIPL thinning was found in both ipsilateral and contralateral subgroups (both P < 0.001). No significant changes were seen in RNFL thickness, INL thickness, retinal microvasculature density, CVV, and CVI. 

Figure 3.

 

OCT/OCTA metrics changes for overall eyes and in ipsilateral eyes or contralateral eyes. (A) Changes before and 4 days after stenting. (B) Changes before and 3 months after stenting. Repeated-measures analysis of variance was conducted for comparison. MD, mean difference.

Figure 4 shows OCT/OCTA metrics changes in the four quadrants. Four days after stenting, RNFL thickness increased in the inferior and nasal quadrants (both P < 0.05); a similar change was seen in the SVC density (both P < 0.05). INL thickness increased in the inferior quadrant (P = 0.046). CVV increased in all four quadrants (P < 0.05). Contrarily, GCIPL thickness decreased in the superior, temporal, and nasal (all P < 0.05) quadrants. Three months after stenting, GCIPL thinning was found in the four quadrants (all P < 0.001). No significant differences were shown in RNFL thickness, SVC density, and CVV (P > 0.05). DVC density and CVI in superior quadrants increased significantly 3 months after stenting (P = 0.002 and P = 0.027, respectively). 

Figure 4.

 

Comparison of OCT/OCTA metric in four quadrants. (A) Changes before and 4 days after stenting. (B) Changes before and 3 months after stenting. Repeated-measures analysis of variance was conducted for comparison.

Supplementary Figures S1 and S2 show the changes in OCT/OCTA metrics in asymptomatic and symptomatic subgroups to exclude the effect of cerebral infarction. When comparing the OCT/OCTA metrics before and 4 days after stenting, RNFL, INL, SVC, CVV, and CVI increased while GCIPL thickness reduced (P < 0.05). Thinning in GCIPL thickness at 3 months after stenting was found in both asymptomatic and symptomatic patients (P < 0.001). 

Discussion

We used SS-OCT/OCTA, a noninvasive modality, to evaluate retinal and choroidal changes before and after carotid stenting in patients with CAS. The use of SS-OCT/OCTA enables a fast-scanning speed, permitting denser scan patterns and larger scans. Moreover, the tool operates with a longer wavelength, facilitating enhanced light penetration through the retina and choroid, which may help resolve the controversies surrounding the effect of carotid stenting on the eye. 

Structural retinal thickness reports in CAS have been inconsistent; some reports showed retinal thicknesses were thinner in patients with CAS compared to controls,^18,19 while some studies did not find any significant difference in the retinal structural thicknesses when CAS was compared to controls.^20,21 Here, we showed that RNFL thickness in contralateral eyes was thinner compared to ipsilateral eyes. Most of our patients had cerebral infarction that was ipsilateral to the location of CAS. Accumulating studies have shown that cerebral infarction results in RNFL thinning (transneuronal retrograde degeneration). Here we suggest that RNFL thinning in contralateral eyes compared to ipsilateral eyes may be due to the presence of cerebral infarction, which may deteriorate the thinning of the RNFL layer. Contrarily, GCIPL thickness was thinner in ipsilateral eyes compared to contralateral eyes. It is suggested that thinning of the GCIPL occurs before retinal ganglion cell death and loss, suggesting that GCIPL thickness may be more informative and sensitive to ischemia in the retina. 

The effect of carotid revascularization on the retina has been widely studied recently. Previous reports^22–24 did not find any significant changes in the retinal thicknesses pre- and postsurgery. Four days after stenting, we detected RNFL thickness thickened in contralateral eyes. Importantly, there was thinning of the GCIPL thickness in the ipsilateral and contralateral eyes 4 days and 3 months after carotid stenting. 

Here, we showed that SVC and DVC densities were lower in ipsilateral eyes compared to contralateral eyes. The lower retinal microvascular densities in ipsilateral eyes could be explained by reduced blood flow of the central retinal artery resulting from reduced ophthalmic artery blood flow due to the stenosed carotid artery. Our result is similar to our previous study,²⁵ which showed reduced retinal microvascular density in ipsilateral eyes compared to contralateral eyes. Using the color Doppler imaging tool, a study showed significant improvement in the ocular blood flow after carotid revascularization.²⁶ Previous OCTA reports^27,28 did not find a significant difference in the SVC density pre- and postoperation in ipsilateral eyes. Four days after carotid stenting, SVC density measurements increased in the overall and ipsilateral eyes compared to the baseline data. The SVC is the entry point of blood flow into the retina and is suggested to be sensitive to ischemic changes in the retina²⁵; importantly, this microvascular plexus is responsible for the arterial circulation of the retina.^29,30 We suggest that carotid stenting did result in increased SVC density during the acute postoperative phase. In terms of a longer postoperative period, we did not find any significant differences in the retinal microvasculature in the ipsilateral and contralateral eyes. 

The SVC is responsible for the metabolism of the retinal ganglion cells (GCIPL).³¹ Accumulating studies^32–34 have shown that changes in the SVC reflect the GCIPL and vice versa. Here, we showed a significant increase in the SVC density and a significant GCIPL decrease 4 days after stenting in ipsilateral eyes. A cerebral imaging report suggested that cerebral vessels are dilated during carotid stenosis and take a longer time to return to their normal state/size after revascularization.³⁵ Given the retina microvasculature and cerebral vasculature have similar anatomy, physiology, and embryology features,³⁶ we suggest that similar changes occur in the retina. It is plausible to suggest that revascularization may have a rapid impact on the superficial retinal blood flow (density) on a short-term basis and may take a longer time to influence the retinal ganglion cells (GCIPL). 

The choroid is the most vascularized tissue in the eye and is primarily supplied by the posterior ciliary arteries. There has been growing interest in exploring the changes in choroidal perfusion in patients with carotid artery stenosis and the effect of surgery on its perfusion. A recent study¹¹ used a semiautomated algorithm to analyze the effect of CEA on the choroid in patients with carotid artery stenosis; the authors showed choriocapillaris perfusion was increased in the ipsilateral eyes compared to contralateral eyes while choroidal thickness was lower in ipsilateral eyes compared to the contralateral eyes. Here, our baseline analysis showed that ipsilateral eyes had lower CVV compared to contralateral eyes while no significant differences were in the CVI. These findings are in line with our previous report.⁹ 

Four days after carotid stenting, we did show a marked increase in CVV measurement in ipsilateral and contralateral eyes. Previous reports that investigated acute choroidal thickness changes during the immediate postoperative period have been inconsistent. In line with our findings, two studies^9,37 showed that choroidal thickness increased within a week after carotid artery stenting, while two recent studies^11,38 showed choroidal thickness increased within a week after CEA. The authors suggested that as a result of reperfusion of the internal carotid artery, the blood volume of the choroid increases immediately. Contrarily, a previous study³⁹ did not show any significant change in the choroidal volume and choroidal thickness on the second day after CEA. Our study suggests that carotid artery revascularization following stenting resulted in increased choroidal perfusion in the acute postoperative period increasing CVV. The long-term effects of carotid revascularization were found in several studies in terms of choroidal circulation. Previous reports^24,39,40 showed an increase in choroidal thickness 3 months after carotid revascularization. Interestingly, two studies^39,40 showed contralateral eyes had increased choroidal thickness 3 months after surgery while significant improvement was not seen in ipsilateral eyes. It has been shown that the circle of Willis artery redistributing blood to the contralateral brain and eye has a positive effect on the contralateral cerebral and ocular perfusion (redistribution theory). However, our study did not show any significant change in the CVV in the ipsilateral and contralateral eyes 3 months after carotid stenting. 

The choriocapillaris (CC) is the innermost layer of the choroid, responsible for the metabolism of the retinal pigment epithelium and photoreceptors.⁴¹ The CC represents a small portion of the choroid and serves as the capillary bed in the choroid. Importantly, the CC represents the terminal capillaries of the choroid and connects the choroidal arterial and venous circulations, making it sensitive to changes in the choroidal blood flow.⁴¹ In our previous study,⁹ we assessed the CC perfusion (length of perfused choriocapillaris per unit area in square millimeters, mm²) and did not observe any significant difference after stenting. Pierro et al.⁴² also analyzed the choroidal vasculature using an external algorithm to assess the CC perfusion; the authors did not observe any significant change between the ipsilateral and contralateral eyes after CEA. Durusoy et al.⁴³ showed an increase in CVI on the first day and first month after stenting in patients with severe CAS. However, their CVI measurements were manual and based on a single B-scan centered on the fovea. We did not see any change in the CVI on either ipsilateral or contralateral eyes at baseline, which is in line with our previous report⁴⁴; similarly, 4 days after carotid stenting and 3 months after stenting, no significant difference was seen in the CVI. The CVI is the ratio of the choroidal vascular luminal volume to the total choroidal volume; it is suggested that the CVI changes when the CVV changes without a proportionate change in the choroidal stroma. Despite changes in choroidal thickness, CVI may remain constant over time and across macular regions. Although we suggest that CVV and choroidal stroma may change proportionately after stenting, we suggest that CVI may remain unchanged after stenting. 

The strengths of this study are relatively large samples of patients with confirmed CAS, detailed OCT/OCTA examination before stenting, and follow-up at 4 days and 3 months after stenting. We also quantitatively analyzed the retinal structural and microvascular change sector-wise to explore which quadrant is sensitive to ischemic changes in the retina. Our findings may suggest that the inferior quadrants are more sensitive to ischemic changes in the acute poststenting phase. 

Some limitations should be acknowledged in our study. First, the study was conducted over a relatively short period, with an average follow-up of 4 days and 3 months. As a result, it is unknown whether the improvements in the retina and choroid will last over a longer period. A longer follow-up of this cohort would be necessary to determine whether carotid artery stenting improves ocular perfusion in the long run. In addition, the patients with retinal artery or vein occlusion and other severe ophthalmic diseases were excluded from data analysis, which might induce selection bias. In the current study, we did not investigate the correlation between the changes of retina/choroid and DSA parameters, such as stenotic degree, which may limit the interpretation of the retinal and choroid changes. The ocular blood flow rate is one of the most important parameters in evaluating ocular perfusion following carotid stenting, and it can be measured using color Doppler or magnetic resonance imaging. In studying the choroid, it would be useful to conduct velocimetry measurements on the choroidal vessels. It would be helpful to investigate the impact of carotid revascularization on functional outcomes such as visual acuity and cognition in future studies. In addition, we did not assess the axial length of our study participants. It is widely known that axial length influences measurements of OCT/OCTA metrics. Although longitudinal design may alleviate the influence of axial length, the finding in our current study may be restricted, and future studies should include the assessment of axial length in the study design. 

In conclusion, our quantitative OCT/OCTA investigation showed that 4 days after carotid stenting, RNFL thickness, SVC density, and CVV increased while GCIPL thickness reduced. We also showed GCIPL thinning 3 months after stenting. The long-term effect of carotid stenting on the retina and choroid should be explored by longer follow-up. 

Acknowledgments

Supported by the National Natural Science Foundation of China (82071320, 8601022), Post Doctor Research Project, West China Hospital, Sichuan University (2021HXBH081), Sichuan Science and Technology Program (2023NSFSC1558), Medical-Engineering Integration Interdisciplinary Talent Training Fund Project of West China Hospital, Sichuan University, and University of Electronic Science and Technology of China (HXDZ22011/ZYGX2022YGRH017). 

The data that support the findings of this study are available on request from the corresponding author. 

Disclosure: L. Cao, None; J. Wu, None; H. Wang, None; W.R. Kwapong, None; Y. Yan, None; J. Wan, None; P. Wang, None; G. Liu, None; R. Wang, None; F. Hu, None; L. Feng, None; B. Wu, None