Article Text
PDF
Abstract
Aims To describe the vascular abnormalities in patients affected by Stargardt disease (STGD1) by means of optical coherence tomography angiography (OCT-A).
Methods Cross-sectional case series, with the following inclusion criteria: diagnosis of STGD1, clear ocular media, and stable fixation. Patients underwent best-corrected visual acuity (BCVA), biomicroscopy, applanation tonometry, short-wavelength fundus autofluorescence (SW-FAF) (HRA Heidelberg, Germany), 3×3 Swept Source OCT-A (Topcon Corporation, Japan). Foveal avascular zone (FAZ) area was manually outlined and removed from the vessel density analysis (ImageJ). Main outcome was vessel density assessment in the superficial capillary plexus (SCP), in the deep capillary plexus (DCP), and in the choriocapillaris (CC) of patients with STGD1.
Results Nineteen patients (36 eyes) were recruited for the study (10 females, 52.6%). Mean age was 33±5.7 years and mean BCVA was 0.6±0.3 logarithm of the minimum angle of resolution. Thirty-six healthy age-matched subjects (one eye for each patient) acted as a control group. Qualitative analysis of OCT-A revealed areas of reduced vascular density in superficial and DCPs. CC showed focal defects partially corresponding to the flecks on SW-FAF imaging. Quantitative analysis of OCT-A disclosed a statistically significant difference in the density of the SCP (0.302±0.062 vs 0.365±0.042; p=0.0002) and the DCP (0.303±0.081 vs 0.399±0.045; p<0.001) compared with controls. To analyse CC, patients with STGD1 were divided into two groups, according to the presence of chorioretinal atrophy. Patients with atrophy showed significantly lower CC density compared with controls (p=0.0003) and patients without atrophy (p=0.001). Patients with STGD1 showed a larger FAZ at the SCP level compared with controls (p=0.012).
Conclusions Vascular impairment in patients affected by STGD1 is concentrated in superficial and the deep retinal plexuses. Patients with atrophic changes have a greater reduction in CC density compared with controls (‘dark atrophy’). Morphological vascular evaluation may become an important step for predicting STGD1 treatment outcomes.
- Retina
- Diagnostic tests/Investigation
- Dystrophy
- Macula
Statistics from Altmetric.com
Introduction
Stargardt disease (STGD1) is the most common form of inherited juvenile macular dystrophy and is caused by a mutation in the ATP binding cassette transporter 4 (ABCA4) gene, which is inherited with an autosomal recessive pattern. Mean age at onset is 15.2 years and its prevalence is estimated at 1 in 10 000.1 Fundus examination shows a beaten-bronze or a bull's eye macular appearance and characteristic deep yellowish-white fish-shaped flecks in the macular and perimacular region. The end-stage fundus aspect is characterised by extensive retinal pigment epithelium (RPE) and choriocapillaris (CC) atrophy, with resorbed flecks and sparse pigmentation.2
STGD1 has been extensively studied by multimodal imaging. Fundus autofluorescence (FAF) shows hyperautofluorescence corresponding to the yellowish-white fundus flecks and hypoautofluorescence at the level of the RPE atrophy.3–6 Fluorescein angiography typically reveals blockage of choroidal fluorescence, referred to as the dark choroid sign, due to lipofuscin accumulation.3–6 Optical coherence tomography (OCT) reveals changes in the outer nuclear layer, and photoreceptor loss, RPE abnormalities and general thinning of the centre of the retina.6 ,7 Little is known about the degree to which retinal and choroidal vasculature is involved in STGD1.
Full-spectrum amplitude-decorrelation angiography associated with OCT angiography (OCT-A) enables vascular abnormalities to be detected at the level of retinal and CC plexuses, which can help us understand the pathogenesis of the disease.8–11 The aim of the study is to describe the vascular abnormalities in patients affected by STGD1, calculating the density of the superficial and deep vascular networks, and of the CC in the macular region by means of OCT-A, along with the area of the foveal avascular zone (FAZ).
Material and methods
This is an observational cross-sectional study. A consecutive series of patients affected by STGD1 referring to the Department of Ophthalmology of San Raffaele Hospital, in Milan, were enrolled for the study between October 2015 and June 2016. Written informed consent was obtained from all subjects. The protocol was approved by the institutional review board of San Raffaele Hospital and the procedures followed the tenets of the Declaration of Helsinki. Inclusion criteria were the diagnosis of STGD1, along with clear media and stable fixation, so as to allow an adequate OCT-A examination. The exclusion criterion was the presence of any other ocular disorder. Thirty-six healthy age-matched controls (36 eyes included in the analysis, one eye for each subject) without any ocular or systemic disease were included in the analysis.
Each patient underwent a complete ophthalmic examination, including best-corrected visual acuity (BCVA) expressed in logarithm of the minimum angle of resolution, biomicroscopy, applanation tonometry, short-wavelength FAF (SW-FAF) (Spectralis, HRA Heidelberg, Heidelberg, Germany), spectral domain OCT, and OCT-A. The classification of STGD1 stages was performed according to the classification of Fishman et al5 as follows: phenotype I (small atrophic-appearing foveal lesions and localised perifoveal yellowish-white flecks), phenotype II (numerous yellowish-white fundus lesions throughout the posterior pole) and phenotype III (extensive atrophic-appearing RPE changes). All patients also underwent molecular analyses for gene ABCA4.
Three by three OCT-A centred on the macula was performed by means of a Swept Source DRI OCT Triton (Topcon Corporation, Japan). This instrument has an A-scan rate of 100 000 scans per second, wavelength of scanning light centred on 1050 nm and in-depth resolution of 2.6 μm (digital). Each OCT-A scan contains 256 B-scans (each B-scan contains 256 A-scans). To image the motion of scattering particles (erythrocytes), four sequential OCT raster scans were repeated at the same location (assisted by eye tracking). Images were analysed with the Topcon full-spectrum amplitude decorrelation angiography algorithm. Automated segmentation of full-thickness retinal scans into the superficial (SCP) and deep (DCP) inner retinal vascular plexuses, outer avascular retina, and CC was performed. Manual adjustment of the segmentation was performed in cases of severe alteration of macular cytoarchitecture. The superficial inner retina contains the vasculature of the retinal nerve fibre layer and ganglion cell layer; the deep inner retina contains the vasculature of the inner plexiform layer, the inner nuclear layer and the outer plexiform layer (OPL).
All 3×3 OCT-A images were exported from the system as a Joint Photographic Experts Group (JPEG) file into the National Institutes of Health ImageJ 1.50 (National Institutes of Health, Bethesda, Maryland, USA) software for the analysis. The FAZ area was manually outlined through the free-hand selection tool, and its dimension was expressed as squared millimetres, as previously described.12 To calculate capillary vessel density, each image was converted from 8-bit into red green blue (RGB) colour type; the image was split into the three channels (red, green and blue) and the red channel was chosen as the reference. The adjust threshold tool set to default was applied; the dark-background option was selected. This tool automatically set lower and upper threshold values (arbitrarily chosen as 110–255, respectively, for every image), and segments greyscale images into features of interest and background. Binarized image was converted back to RGB; FAZ area was restored and coloured to pure blue. White pixels were considered as vessel, black pixels as background, and blue pixels were automatically excluded from the analysis (see online supplementary figure S1). Retinal vessel density was defined as the total pixel areas of the retinal vessels in the middle retinal layer of the OCT angiogram divided by the total area between the FAZ and the remaining area, similarly to published studies;13 SCP, DCP and CC of the patients and the controls were analysed with this method. Sets of obtained values were compared with controls for each different layer of segmentation (SCP, DCP or CC) with the GraphPad Prism software 5.0 (GraphPad software, Inc., San Diego, California, USA). Descriptive statistics for demographics and main clinical records, comparative analysis (Student's t test analysis and one-way analysis of variance (ANOVA) for independent samples) were also performed, and qualitative descriptions of the imaging findings. Tukey correction has been used for post-hoc analysis to find means that are significantly different from each other. The dataset was analysed for normal distribution by means of the d'Agostino-Pearson test. The chosen level of statistical significance was p<0.05.
Supplementary figure
Optical coherence tomography angiography (OCT-A) image processing for quantitative analysis.
Results
Thirty-six eyes of 19 consecutive patients affected by STGD1, confirmed by ABCA4 analysis, were recruited for the study. The demographic and clinical characteristics of the patients and the controls are listed in table 1.
Demographic characteristics of patients with Stargardt dystrophy and controls
Twelve of 36 eyes (33.3%) did not show any atrophic change at the posterior pole; 20 of 36 eyes (55.6%) showed patchy RPE loss, but not extensive atrophy, while the other 4 eyes had advanced macular atrophy (11.1%).
In comparison with the vascular pattern detectable in the control group, OCT-A anomalies were identified in all eyes (100%). In particular, the SCP showed a vascular rarefaction of the normal vascular density, which was especially visible in the temporal parafoveal sector of the macula (figure 1). The OCT-A assessment of the DCP revealed an increased intervascular space with progressive capillary rarefaction within the whole macular area examined (figure 2), suggesting a more complex vascular impairment characterised by a partially ischaemic DCP pattern. The analysis of the CC showed vascular rarefaction characterised by reduced capillary density, which was more evident in the patients affected by chorioretinal atrophy, where large choroidal vessels were well recognisable (figure 3). The most remarkable finding was the identification of small dark areas on the CC related to the masking effect of the flecks, which appeared in a greater number than in the corresponding SW-FAF imaging.
Vascular network differences between patients and controls in the superficial capillary plexus (SCP). Top left: 3×3 optical coherence tomography angiography (OCT-A) of the SCP of a 43-year-old patient affected by Stargardt dystrophy (STGD1); top right: 3×3 OCT-A of the SCP of a healthy 47-year-old man. Rarefaction of the perimacular vascular network is well recognizable in the patient with STGD1 (asterisc). Bottom left: box plot showing a significant difference in the SCP density (p=0.0002) and in the foveal avascular zone (FAZ) area (p=0.012) between patients (A) and controls (B).
Vascular network differences between patients and controls in the deep capillary plexus (DCP). Top left: 3×3 optical coherence tomography angiography (OCT-A) of the DCP of a 25-year-old patient affected by Stargardt dystrophy (STGD1); top right: 3×3 OCT-A of the DCP of a healthy 30-year-old man. Bottom left: box plot showing a significant difference in the DCP density (p<0.001) between patients (A) and controls (B). Bottom right: box plot showing a non-significant difference in the foveal avascular zone (FAZ) area (p=0.2).
Vascular network differences between patients and controls at the choriocapillary (CC) level. Top left: 3×3 optical coherence tomography angiography (OCT-A) of the CC of a patient affected by Stargardt dystrophy (STGD1) with no macular atrophy, as confirmed by colour fundus photograph (bottom left). Top centre: 3×3 OCT-A of the CC of a patient affected by STGD1 with macular atrophy at fundus image (bottom centre), showing severe loss of CC (arrowhead); top right: 3×3 OCT-A of the CC of a healthy 49-year-old control. Bottom right: box plot showing a non-significant difference between patients without atrophy (A1) and controls (B) (p=0.9), a significant difference between patients with atrophy (A2) and controls (B) (p=0.0003), and between patients with atrophy (A2) and those without atrophy (A1) (p=0.001).
Quantitative analysis of OCT-A centred on the macular area disclosed a statistically significant difference in the mean density of the SCP (p=0.0002) and of the DCP (p<0.001), comparing the patients and the control group (table 2; figures 1 and 2). The CC disclosed a little difference taking the STGD1 group of patients all together (p=0.02). Patients with chorioretinal atrophy were divided into patients without chorioretinal atrophy on the bases of colour fundus photograph and SW-FAF; the two groups were compared separately with controls by means of ANOVA. Patients without chorioretinal atrophy did not show any difference from controls (p=0.9). On the contrary, patients with atrophy had significantly lower CC density with respect to controls (p=0.0003) and to those patients without atrophy (p=0.001) (table 2; figure 3).
Quantitative comparison of the macular vascular density and the foveal avascular zone between patients with Stargardt dystrophy and controls
The FAZ area was measured at SCP and DCP level in patients and controls; the mean FAZ turned out to be significantly larger in patients with STGD1 considering the SCP (p=0.012), while it was not different at the level of the DCP (p=0.2) (table 2; figures 1 and 2).
Discussion
The complex pathophysiology of STGD1 involves the progressive accumulation of lipofuscin, the main by-product of the photoreceptor visual cycle, in the RPE during the process of disk shedding. Lipofuscin and its components, especially N-retinylidene-N-retinylethanolamine (A2E), the ultimate product of condensation of all-trans-retinal and N-retinylidene-PE, are toxic to epithelial and neuronal cells, with consequent RPE and secondary photoreceptor degeneration.14
Little is known about the vascular changes related to STGD1. The assessment of vascular changes in retinal dystrophies is useful in trying to understand disease pathophysiology and also for both clinical staging and the relative prognostic implications. With this in mind, we evaluated the macular vascular plexuses in patients with genetically confirmed STGD1.
Our results indicate that OCT-A can provide interesting insights regarding the changes in the vascular patterns of eyes affected by STGD1. Interestingly, all three layers analysed appear to be affected. The eyes displaying atrophic changes at the posterior pole revealed a greater decrease in vascular density at the level of CC, with hyper-reflective response probably due to the absence of RPE and the direct visualisation of the underlying large choroidal vessels (figure 3). At the same time, patients with STGD1 disclosed a larger FAZ compared with controls that reached a statistically meaningful level only at the SCP. These findings suggest a complex vascular impairment, with ischaemic damage to the full-thickness neuroretinal layer in patients affected by early and moderate forms of STGD1. Oxygen is supplied to the photoreceptor by the CC and the DCP.15 Nevertheless, it has been reported that the photoreceptor axon terminals, located in the OPL, are full of oxygen-dependent mitochondria and may rely more heavily on the DCP for oxygenation than the inner and outer photoreceptor segments, which are closer to the underlying CC.16
CC atrophy and RPE loss are strictly related, since these two structures share a close relationship, RPE releasing many cytokines, which in turn stimulate the CC trophism.17 Extensive loss of choriocapillary associated with an intact choroid has been described in aggressive cases of the disease.18 This histopathological condition explains the hypocyanescence on indocyanine green angiography (‘dark atrophy’) within the area of macular atrophy found in 92% of patients with STGD1.19 Our results agree with previously published data, showing that the CC layer is mostly affected in those patients with greater extent of atrophy, while it is partially spared in early stages of the disease.
The considerable vascular deficiency identified on OCT-A at the level of the SCP and DCP may be extremely important in ensuring adequate trophism of the photoreceptors, and may directly influence their progressive degeneration over the course of the disease. The therapeutic implications of this finding are potentially considerable. First of all, in principle preserving the integrity of the DCP's vascular density may improve the oxygenation of the photoreceptors, ensuring their survival and functions. Second, the potential application of gene therapy for STGD1 could be guided by the vascular pattern, as visualised on OCT-A. Thus, eyes with a severely damaged vascular network might not be optimal candidates for this approach. Furthermore, the gene therapy could be tailored to match the blood supply (dose, amount of fluid, location of injection).20–21
We acknowledge that the present study has a number of limitations, including the number of patients, the absence of a longitudinal follow-up, and the subjective classification of the type of the disease. Moreover, signal strength index could be significantly lower in patients with STGD1 compared with healthy subjects and captured images could suffer from a number of motion artefacts due to fixation instability. This technical limitation may have negatively influenced the interpretation of the images included in the study. The OCT-A apparatus used, which is in fact a prototype and has seldom been employed in clinical practice, itself constitutes a further drawback, potentially providing image artefacts. In view of this, the present research should be regarded as the basis for further analyses of the vascular changes related to STGD1 as visualised on OCT-A.
In essence, the present study maintains that OCT-A reveals an abnormal vascular pattern, especially at the level of the CC in advanced forms of the disease. Further studies are warranted to correlate vascular anomalies with functional and therapeutic response in patients affected by STGD1.
References
Footnotes
Contributors All the authors contributed to the conception or design of the work, the acquisition, analysis and interpretation of data, drafting the work, revising it critically for important intellectual content and gave final approval of the version to be published.
Competing interests FB: consultant for Allergan Inc (Irvine, California, USA), Novartis (Basel, Switzerland), Farmila-Thea (Clermont-Ferrand, France), Bayer Shering-Pharma (Berlin, Germany), Alcon (Fort Worth, Texas, USA), Bausch and Lomb (Rochester, New York, USA), Genentech (San Francisco, California, USA), AlimeraSciences (Alpharetta, Georgia, USA), Sanofi-Aventis (Paris, France), Thrombogenics (Heverlee, Belgium), Hoffmann-La-Roche (Basel, Switzerland), NovagaliPharma (Évry, France).
Ethics approval Institutional review board of San Raffaele Hospital.
Provenance and peer review Not commissioned; externally peer reviewed.