Article Text

Download PDFPDF

Optical coherence tomography angiography characteristics of polypoidal choroidal vasculopathy
  1. Mayer Srour,
  2. Giuseppe Querques,
  3. Oudy Semoun,
  4. Ala El Ameen,
  5. Alexandra Miere,
  6. Anne Sikorav,
  7. Olivia Zambrowski,
  8. Eric H Souied
  1. Department of Ophthalmology, Centre Hospitalier Intercommunal de Créteil, Université de Paris Est Créteil, Créteil, France
  1. Correspondence to Professor Eric H Souied, Department of Ophthalmology, Centre Hospitalier Intercommunal de Créteil, Université de Paris Est Créteil, 40 Avenue de Verdun, Créteil 94000, France; eric.souied{at}chicreteil.fr

Abstract

Purpose To analyse the morphological characteristics of polypoidal choroidal vasculopathy (PCV) on optical coherence tomography angiography (OCT-A).

Methods Prospective study with consecutive patients affected with PCV were included. All patients underwent a complete ophthalmological examination including fundus photography, fluorescein angiography, indocyanine green angiography, spectral-domain OCT and OCT-A.

Results Twelve eyes of 12 patients (mean age 72.6±10.5 years; 4 men and 8 women) were included for analysis. In all eyes (12/12) the segmentation of the choriocapillaris layer on OCT-A revealed the branching vascular network (BVN) as a hyperflow lesion. OCT-A segmentation of the choriocapillaris layer in correspondence of the polypoidal lesion showed in 3/12 eyes (25%) a hyperflow round structure, surrounded by a hypointense halo, and in 9/12 eyes (75%) a hypoflow round structure.

Conclusions The OCT-A is a non-invasive imaging modality allowing the visualisation of different structures in PCV. The BVN is constantly clearly detected. The hypoflow round structure appearance of the polyp in OCT-A, is probably due to an unusual blood flow inside the polypoidal lesions, contrasting with the BVN. Further improvement in OCT-A knowledge will provide information on the specificity of the different intensity characteristics in PCV.

  • Imaging
  • Neovascularisation
  • Retina

Statistics from Altmetric.com

Introduction

Age-related macular degeneration1 (AMD) is the first cause of blindness after 50 years in the Western world. AMD is a phenotypically heterogeneous disease, including atrophic and exudative forms. The exudative form is characterised by an abnormal choroidal neovascularisation (CNV) within the macula.2 Polypoidal choroidal vasculopathy (PCV) is an acquired, abnormal choroidal vasculopathy, distinct from typical CNV.3 ,4 The first PCV description was in 1982 by Yannuzzi.5

PCV is characterised by polypoidal dilations and choroidal branching vascular networks (BVNs) observed on indocyanine green angiography (ICGA).3–7 Optical coherence tomography (OCT) is a very useful tool for diagnosis of PCV, as polypoidal dilations are characterised by dome-like elevations of retinal pigment epithelium (RPE) and moderate internal reflectivity.8–11 BVNs appear on OCT images as two highly reflective lines12 (‘double layer sign’).

OCT angiography (OCT-A) is a novel and non-invasive imaging tool and en face derived technique13 ,14 that allows the visualisation of retinal microvasculature by detecting intravascular blood flow, using a split-spectrum amplitude-decorrelation angiography (SSADA) algorithm, without any dye injection. It allows visualising the dynamic motion of erythrocytes using sequential OCT cross-sectional scans and a detailed assessment of the retinal, as well as of the choroidal circulation.

In this study, our purpose was to describe imaging features of PCV and adjacent structures using OCT-A.

Methods

In this prospective study, consecutive patients presenting at Creteil University Eye Clinic with PCV were included, from November 2014 to April 2015. This study was driven in accordance with the Declaration of Helsinki, current French legislation and with approval of our local ethics committee.

For all patients, PCV was diagnosed by ICGA showing the typical BVN and polypoidal dilations3–7 and was ascertained by a retinal specialist (GQ or EHS). Patients were excluded if affected with any other macular disorders such as high myopia (>8 dioptres), presence of angioid streaks or intraocular inflammation.

For each patient, a complete ophthalmological examination was performed as part of their routine clinical work-up. It included a best-corrected visual acuity (BCVA) measure using Early Treatment Diabetic Retinopathy Study (ETDRS) charts, fundus examination, scanning laser ophthalmoscopy fluorescein angiography (FA) and ICGA (Spectralis HRA+OCT; Heidelberg Engineering, Heidelberg, Germany), spectral-domain OCT (Spectralis HRA+OCT; Heidelberg Engineering).

OCT-A was performed in all patients with a scanning area of 3×3 mm, centred on the polypoidal lesions and BVNs, using Optovue RTVue XR Avanti (Optovue, Freemont, California, USA) to obtain amplitude-decorrelation angiography images. This instrument has an A-scan rate of 70 000 scans per second, using a light source centred on 840 nm and a bandwidth of 50 nm. Each OCT-A volume contains 304×304 A-scans with two consecutive B-scans captured at each fixed position before proceeding to the next sampling location. SSADA was used to extract the OCT-A information. Each OCT-A volume is acquired in 3 s and two orthogonal OCT-A volumes were acquired in order to perform motion correction to minimise motion artefacts arising from microsaccades and fixation changes. Angiography information displayed is the average of the decorrelation values when viewed perpendicularly through the thickness being evaluated.

For this study, we have used the software embedded in the machine. In our study, we have used only automatic segmentation without any adjustment of the layer boundaries. Automatic segmentation distinguishes four retinal sections, allowing the separate analysis of blood flow in each section.

The inner retinal layer's limits were defined from the internal limiting membrane to the outer boundary of the outer plexiform layer (OPL). Thus, the inner retina layer contained all of the normal retinal vasculature. The outer retinal layer's boundaries were defined from the outer OPL to the Bruch's membrane (BM). The choroidal layer was defined as any structure located below BM.

The OCT-A characteristics were analysed and compared with ICGA. Associated neovascular lesions possibly resulting from AMD, such as occult (type 1) or classic (type 2) CNV, were noted and described when present. The ability to detect polypoidal dilations and BVN on OCT-A was assessed by two readers (MS and GQ). Disagreement between readers was resolved by open adjudication.

Results

Twelve eyes from 12 consecutive patients with PCV were included. Patients were four men and eight women, aged from 57 to 89 years (mean 72.6±10.5 years). The mean time since diagnosis was 16.8±13.8 months. The localisation of the PCV was subfoveal or juxtafoveal in 6/12 cases, and peripapillary in 6/12 cases. Eyes were naive to treatment in 2/12 cases. In the remaining 10/12 cases, eyes were previously treated with intravitreal anti-vascular endothelial growth factor (VEGF) agents (mean 12.2±13.8 intravitreal injections) or with photodynamic therapy (PDT) (mean 1.8±0.8 PDT). For all combined groups, mean BCVA was 0.32±0.24 LogMAR. In one eye, a type 2 neovascularisation was detected in association with PCV.

In all eyes, the segmentation of the choriocapillaris layer on OCT-A revealed the BVN as a hyperflow lesion (figures 13). Comparison between OCT-A pictures and mid/late ICGA phases revealed that hyperflow lesion in the choriocapillaris layers corresponded topographically to the BVN visualised on ICGA (figures 13).

Figure 1

Multimodal imaging of the left treatment-naive eye of a 67-year-old woman with polypoidal choroidal vasculopathy associated with classic (type 2) choroidal neovascularisation (CNV). Fundus photograph (top left), fluorescein angiography (top middle), indocyanine green angiography (top right) and spectral-domain optical coherence tomography (OCT) (middle left) showing the polypoidal lesion and corresponding abnormal choroidal vascular network associated with classic CNV. (Lower left, lower right) OCT angiography image of the choriocapillaris segmentation (double red line) with corresponding OCT B-scan reveal the branching vascular network as a hyperflow lesion (grey arrow) and the polypoidal lesion as hypoflow round structure (white arrow). The classic CNV appears as a hyperflow structure (black arrow).

Figure 2

Multimodal imaging of the left eye of a 67-year-old Asian woman with polypoidal choroidal vasculopathy after treatment (13 intravitreal anti-VEGF injections). Fundus photograph (top left), fluorescein angiography (top middle), indocyanine green angiography (ICGA) (top right) and spectral-domain optical coherence tomography (OCT) (middle left) showing the polypoidal lesion and corresponding abnormal choroidal vascular network. Note on phase ICGA (top right) three hyperfluorescent round structures (white stars) corresponding to polypoidal lesions. OCT angiography image in choriocapillaris segmentation (double red line) with corresponding B-scan (lower left, lower right) reveal the branching vascular network as a hyperflow lesion (black arrow) and the three polypoidal lesions as hyperflow round structure, surrounded by a hypointense halo (white arrow).

Figure 3

Multimodal imaging of the right eye of a 66-year-old man with polypoidal choroidal vasculopathy after treatment (12 intravitreal anti-VEGF injections). Multicolour imaging (top left), fluorescein angiography (top middle), indocyanine green angiography (top right) and spectral-domain optical coherence tomography (OCT) (middle left) showing the polypoidal lesion and corresponding abnormal choroidal vascular network. OCT angiography image in choriocapillaris segmentation (double red line) with corresponding B-scan (lower left, lower right) reveal the branching vascular network as a hyperflow lesion (black arrow) and the polypoidal lesion as hyperflow round structure, surrounded by a hypointense halo (white arrow).

OCT-A segmentation of the choriocapillaris layer in correspondence of the polypoidal lesion showed a hyperflow round structure surrounded by a hypointense halo in 3/12 eyes (25%), or a hypoflow round structure in 9/12 eyes (75%). The comparison between OCT-A pictures and mid or late phases of ICGA revealed that either the hyperflow round structure, surrounded by a hypointense halo, or the hypoflow round structure, both corresponded topographically to the borders of the polypoidal lesions as visualised on ICGA (figures 13).

There was no difference between previously treated and treatment-naive eyes concerning the OCT-A characteristics in this series. The two treatment-naive eyes appeared either as a hyperflow round structure, surrounded by a hypointense halo in one case, or as a hypoflow round structure in the other case.

Discussion

In this series, we analysed the morphological characteristics of PCV on OCT-A. OCT-A is a new, non-invasive imaging, using the SSADA,13 ,14 that allows the visualisation of retinal microvasculature by detecting intravascular blood flow and the morphology of vessels in the inner/outer retina and in the choroid without any dye injection. Recent OCT-A studies presented detailed images of CNV,15 of pathological vascular changes in diabetic retinopathy,16 of retinal vascular plexus in telangiectasia type 217 and of microvascular networks in glaucomatous optics discs.18

In these patients, we demonstrated that OCT-A could visualise the different structures in PCV. In the segmentation of the choriocapillaris, OCT-A constantly showed the BVN which appeared as a hyperflow lesion and the polypoidal lesions appearing either as hyperflow round structure surrounded by a hypointense halo, or as hypoflow round structure. Comparison between OCT-A and ICGA pictures revealed that OCT-A characteristics of PCV corresponded topographically to the BVN and the polypoidal lesions as visualised on ICGA.

PCV has been widely studied with FA, ICGA and OCT.3–12 En face imaging of PCV provides an in vivo tool to visualise the pathological features and the choroidal vasculature in PCV without dye injection.19 ,20 In the current study, OCT-A segmentation of choriocapillaris layers, presented the PCV structure with different appearance.

In most cases the polypoidal lesions appeared as hypoflow round structures, this absence of signal does not mean that there is no blood flow, but rather that blood flow is not within the level of detection of the OCT-A device. This could be due to an increase or decrease flow in the polyps, which may be responsible for the non-visualisation of the vascular structure. Choroidal blood flow is known to be higher than retinal blood flow,21 and studies of haemodynamics in PCV suggest that the origin of polypoidal lesions would be from the choroidal vessels.22 The hypothesis of a too high blood flow in the polyp is theoretically possible, but very unlikely because polyps do not fill very rapidly on ICGA.

Thus, another possible explanation for the absence of OCT-A signal could be the presence of turbulent blood flow inside the polypoidal lesions impeding the representation of this flow. In agreement with our findings, the microaneurysms in diabetic retinopathy, which share similitudes structure with polypoidal lesions (turbulent blood flow within the lesion), were not all visualised in OCT-A compared with FA.23

The last explanation, could be that the blood circulates only at the periphery of the aneurysmal dilation, but we did not compare the signal pattern on OCT-A and the dynamics of dye filling in the polypoidal lesions on ICGA.

The various structural OCT-A aspects of polyps in our series can be explained by the effect of anti-VEGF or PDT treatments on blood flow. Miura et al24 have shown that composite Doppler OCT B-scan image shows the presence of blood flow inside the polypoidal lesions and the disappearance or reduction of blood flow after treatment. However, the aim of the current study was simply to describe and thus we did not compare PCV structures in the course of treatment.

Indeed, OCT-A allows the visualisation of retinal microvasculature by detecting intravascular linear blood flow.12 On the other hand, the BVN, which is characterised by linear blood flow, was clearly detected by the SSADA algorithm in the segmentation of the choriocapillaris.

While histopathological studies have indeed demonstrated that the BVN associated to PCV is located in the BM,25 ,26 our visualisation in OCT-A at that precise level is impaired mainly by the projection artefact. It is an important issue in OCT-A technique, recently elucidated by Spaide et al.27 The projection artefact generates a mirror effect, secondary to the existence of hyper-reflective structures such as the RPE. Thus, the visualisation of the branching network in the choriocapillaris segmentation is very likely to involve the projection artefact. These findings give rise to an imperfect, sometimes erroneous correlation between the segmentation seen on the corresponding B-scan and the angioflow image.

Recently, Inoue et al28 published an interesting paper regarding PCV studied by OCT-A. They showed that the BVN signal was relatively homogenous and provided anatomical information comparable to ICGA. Regarding the morphological characteristics of polyps in OCT-A, they demonstrated that polyps were poorly visualised. On one hand, this study demonstrated that cross-sectional OCT-A images provided additional information and revealed that flow signal within the polyps was clear in most of the cases. However, this signal was non-uniform and detected only in a focal area. This suggests that there is differential flow in the polyps. On the other hand, in our study we focused on the morphological features of polyps in OCT-A, which appeared either as a hypoflow round structure, or as a hyperflow round structure, surrounded by a hypointense halo.

The present study has some limitations, mainly related to the relatively limited size of our cohort, and to the current limitations of the OCT-A device. OCT-A provides only 3×3 mm field of view, covering limited areas of the fundus; however, all PCV were easily observed in all cases in our series. Great images acquisition depended of good mydriasis and good patient cooperation. The decorrelation signal is not proportional to the flow velocity, and will be identical above a certain value (this phenomenon is known as ‘saturation limit’). Absence of signal does not mean that there is no blood flow. Thus, there is a possibility of overlooking to a certain degree structures within which blood flow does not reach the level of detection in the device. Furthermore, the possibility of artefact cannot be totally ruled out in this new imaging.

Finally, OCT-A could not reveal the breakdown of blood retinal barrier, which is an important sign of PCV activity.

In conclusion, OCT-A is a non-invasive imaging modality, and in this study we demonstrated that it allows the visualisation of different structures in PCV. The BVNs are clearly constantly visualised as hyperflow lesions, but further improvement in OCT-A knowledge are needed to gather information on the specificity of the different intensity characteristics of polypoidal lesions.

References

View Abstract

Footnotes

  • Contributors Design and conduct of the study: MS, GQ and EHS. Collection, management and analysis: MS, GQ, AEA, OS and EHS; interpretation of the data: MS, GQ and EHS and preparation, review or approval of the manuscript: MS, GQ, AEA, OS, AM, AS, OZ and EHS.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Ethics approval Ethic Committee of Centre Hospitalier Intercommunal de Créteil.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • At a glance
    Keith Barton James Chodosh Jost Jonas