Introduction

Carotid artery stenosis is an important risk factor for ischemic stroke and transient ischemic attacks1. Management of vascular risk factors, antiplatelet therapy and different surgical procedures (carotid endarterectomy, carotid angioplasty and stenting) are therapeutic options available for the management of carotid artery stenosis. Detection of stenosis of the carotid artery is therefore especially important in neurologically asymptomatic patients2,3,4. In current guidelines, the degree of stenosis is an important surrogate measure for stroke risk and indication for intervention. Various imaging technologies such as CT angiography, magnetic resonance angiography and/or duplex ultrasonography are therefore used to evaluate patients with CAS5,6.

Internal carotid artery stenosis can be associated with impaired ocular blood flow and retinal examination is generally performed in internal carotid artery stenosis patients when clinical ocular symptoms such as sudden or progressive visual loss occur7. However, chronic carotid artery stenosis is not necessarily associated with morphological or functional retinal damage7.

Optical coherence tomography angiography (OCT-A) is a novel technology, providing high-resolution images of the retinal vasculature8. This method also enables quantitative evaluation of the retinal blood flow and blood flow in the optic nerve head (ONH)9,10,11.

The aim of the study presented here was to evaluate the retinal and optic nerve head (ONH) perfusion, as measured using optical coherence tomography angiography (OCT-A) in patients with asymptomatic severe internal carotid artery stenosis compared to healthy controls, and to evaluate the impact of carotid endarterectomy.

Methods

Patients

For this prospective study 25 eyes of 25 patients suffering from internal carotid artery stenosis were consecutively enrolled. A control group of 25 eyes of 25 healthy subjects without ocular disease of any sort were also included. 18 patients were planned for surgical treatment and were evaluated before and after surgery. The study followed the tenets of the Declaration of Helsinki and was approved by the Ethics Committee of the University of Muenster, North Rhine Westphalia, Germany. Before imaging, the study protocol was explained in detail and all participants signed an informed consent form.

Inclusion criteria were asymptomatic patients with carotid artery stenosis ≥70% with perioperative risk <3% (Guidelines for the primary prevention of stroke of the American Heart Association/American Stroke Association12), age over 18 years, and a planned surgical treatment.

Exclusion criteria were ocular symptoms (visual loss or visual impairment), media opacities preventing high-quality imaging, vitreoretinal disease or status post vitreoretinal surgery. Patients with neurological diseases, myocardial infarction or strokes were excluded. Patients with diabetes without diabetic retinopathy were included. The comorbidities of the study group and control group are summarized in Table 1 (Table 1).

Table 1 Clinical characteristics of the study groups.
Full size table

Before imaging subjects were also asked to take a rest of about 5 minutes and systemic blood pressure was measured in the left brachial artery at the height of the heart with the subject in an upright sitting position. Subjects with systolic blood pressure (>150 mmHg/<100 mmHg) or diastolic blood pressure (>90 mmHg/ <60 mmHg) were not included.

Surgical Treatment

Surgical treatment was performed in the Department of Vascular Surgery at the University of Muenster Medical Center. Surgery was carried out under general anesthesia. After exposure of the carotid bifurcation and administration of a bolus of heparin, the internal, common and external carotid arteries were clamped. A shunt was placed to maintain perfusion of the intracranial vessels during the procedure and a meticulous removal of the plaque was performed. Depending on the anatomical characteristics of the bifurcation, an eversion endarterectomy with direct suture or a longitudinal endarterectomy with Dacron patch was carried out. All patients were postoperatively monitored in an intermediate care ward for at least 24 hours.

Examination

Before OCT-A imaging, patients underwent a complete ocular examination including refraction, IOP (intraocular pressure) measurement, slit lamp biomicroscopy and funduscopy. OCT-A imaging was performed before and 3–4 days after surgery. OCT-A imaging was performed using the AngioVue device (RTVue XR Avanti with AngioVue, OptovueInc, Fremont, California, USA). This system has an A-scan rate of 70,000 scans/second and the split-spectrum amplitude-decorrelation angiography (SSADA) algorithm was used to generate the angiography data. The OCT-A technology has been described in detail elsewhere9,11. Briefly, repeated OCT scans of a certain area are performed and the OCT images of that area evaluated to identify possible changes. Blood flow in the retinal vessels will result in changes between the successive OCT images, whereas static tissue will show no change11.

OCT-A imaging was performed in the same location by an expert examiner under the same conditions. Imaging of the optic nerve head (ONH) required a 4.5 × 4.5 mm2 scans while macula imaging required a 3.0 × 3.0 mm2 scan. Images of poor quality (lines or gaps arising from poor signal strength or motion artifacts) were excluded from the study. The software automatically segmented the tissue into 4 layers: in the ONH (optic nerve head, vitreous, radial peripapillary capillary (RPC), and choroid) and in the macula (superficial, deep, outer retina and choriocapillaris). After checking the segmentations, the flow density data in the optic nerve head (radial peripapillary capillary (RPC) layer and the macula (superficial and deep retinal OCT angiogram) were then extracted and analyzed.

Data analysis and statistics

Data management was performed using Microsoft Excel 2013. IBM SPSS® Statistics 22 for Windows (IBM Corporation, Somers, NY, USA) was used for statistical analyses. The normality of the data distribution was tested using the Kolmogorov–Smirnov test. After confirmation of the normality assumption data are generally presented as mean ± standard deviation while changes at follow-up compared with baseline were assessed using paired sample t-tests. The two treatment groups were compared using independent Student’s t-tests. All inferential statistics are intended to be exploratory, not confirmatory, and are interpreted accordingly. The global statistical significance level was set to 0.05.

Results

25 patients with CAS (age: 64.56 ± 7.23) and 25 healthy control subjects (age: 64.76 ± 9.81) were prospectively included in the study. There was no statistically significant difference in age between the two groups (p = 0.94). Clinical characteristics of the study group and the control group are summarized in Table 1.

There was no significant difference between the signal strength indices (SSI) in the control group and the study group (SSI of the macula measurements: study group: 69.23 ± 7.57; control group: 69.81 ± 6.27; p = 0.78; SSI of the ONH measurements: study group: 63.78 ± 7.84, control group: 68.18 ± 4.32, p = 0.07).

The flow density (whole en face) in the superficial retinal OCT angiogram of the macula in patients with CAS was significantly lower compared with healthy controls (study group: 48.52 ± 4.46; control group: 51.88 ± 2.70; p = 0.003) (Fig. 1). Significant differences were also found in the ONH. The flow density data in the macula and ONH of the study group and the control group are summarized in Table 2.

Figure 1
figure 1

OCT angiograms of a patient with CAS (Top row) and a healthy control (Bottom row).

Full size image
Table 2 Flow density in the macula and optic nerve head of patients with CAS and of healthy controls.
Full size table

In patients with CAS there were no significant differences in flow density between the ipsilateral and the contralateral eye at baseline (p > 0.05). The flow density (whole en face) in the RPC in the ipsilateral eyes, improved significantly after carotid endarterectomy. The flow density values in the ipsilateral eyes are summarized in Table 3 and the flow density values of the contralateral eyes in Table 4.

Table 3 Flow density values obtained in the indicated regions in the ipsilateral eyes before and after carotid endarterectomy.
Full size table
Table 4 Flow density values obtained in the indicated regions in the contralateral eyes before and after carotid endarterectomy.
Full size table

Discussion

This pilot study is the first to determine reduced flow density in patients with CAS compared with healthy controls using OCT-A. OCT-A is non-invasive and can be performed easily and fast. It enables visualization of blood flow in the retina and ONH without intravenously injected dye. This technology has attracted a great deal of clinical research interest over the last two years and is finding increasing use in clinical practice9. OCTA also enables quantitative analysis of flow density in the retina and optic nerve head and has been assessed in various ocular and systemic diseases9,10,13,14,15,16. The reproducibility of the quantitative analysis of flow density has been evaluated in healthy subjects and in patients with different ocular diseases9,10,14.

Various studies in the literature have evaluated morphological and functional ophthalmological parameters in patients with CAS compared with healthy controls: Sayin et al. found a decreased choroidal thickness in patients with CAS, while in the same study no significant difference was found in the thicknesses of the retinal nerve fiber layer, macula or ganglion cell complex17. Whereas Heßler et al. also failed to find a significantly reduced RNFL thickness in patients with CAS7, a community-based study recently published by Wang et al. does describe reduced RNFL thicknesses in patients with CAS4. In functional tests, Kofoed et al. described significantly reduced and delayed electroretinographic responses in patients with carotid artery stenosis18.

Stenosis of the carotid artery leads to a fall in ocular blood flow7,19,20,21. In our study, patients with CAS showed a reduced flow density in the RPC layer of the ONH and in the superficial retinal OCT-angiogram when compared with healthy controls. In the superficial retinal OCT-angiogram the differences in the parafovea and in the entire evaluated area (whole en face) were significant whereas the difference in the fovea did not reach the significance level. This could be explained by interindividual variation in the area of the FAZ (higher SD in the fovea when compared with parafovea)22. The difference between the two groups in the deep retinal OCT-angiogram was also not significant. However, the analysis of flow density values in the deep retinal OCT-angiogram should be interpreted with caution, since the quantification of flow density in the deep retinal OCT angiogram is more challenging, being affected by projection artefacts, and repeatability was found to be weaker compared with that of the superficial retinal OCT-angiogram in previous studies11,23,24,25.

Most of the patients included in our study had a bilateral CAS. Therefore there were no significant differences between the flow density of the ipsilateral eye and the contralateral eye at baseline. After carotid endarterectomy the flow density improved significantly in the ipsilateral and in the contralateral eye. Carotid revascularization surgery improves cerebral perfusion and has a positive effect on the contralateral cerebral blood flow through the collateral circulation26,27,28. The positive effect of carotid endarterectomy on the ipsilateral retinal blood flow has been reported before21. Lareyre et al. also demonstrated bilaterally increased choroidal thickness using enhanced depth imaging optical coherence tomography (EDI-OCT) in patients with CAS after carotid endarterectomy26. Lareyre et al. hypothesized that EDI OCT could be a potential marker for the assessment of cerebral and/or ocular perfusion after carotid endarterectomy26. An important issue to consider in this context is that collateral pathways through the ophthalmic artery may be recruited to compensate for diminished cerebral blood flow in patients with internal carotid artery stenosis5. Although OCT-A is fast, non-invasive, accurate and reproducible an important limitation of OCT-A that should be mentioned here is the absence of flow direction.

This pilot study is also limited by the small sample size and short follow-up period. Further studies with a larger number of patients and a longer follow-up period are now required to evaluate whether OCT-A could be useful as a potential marker in the diagnosis of CAS and the evaluation of treatment success.

To conclude patients with CAS showed a reduced flow density in the RPC layer (ONH) and in the superficial retinal OCT-angiogram when compared with healthy controls. The flow density improved significantly after surgical treatment (carotid endarterectomy). Measurement of OCT-A and quantitative analyses of flow density could represent a useful, fast, non-invasive and objective approach to diagnosis of CAS and evaluation of treatment success.