Article Text

Download PDFPDF

Characteristics and quantification of vascular changes in macular telangiectasia type 2 on optical coherence tomography angiography
  1. Lavanya Chidambara1,
  2. Santosh G K Gadde1,
  3. Naresh Kumar Yadav1,
  4. Chaitra Jayadev1,
  5. Devanshi Bhanushali1,
  6. Abhishek M Appaji2,
  7. Mukunda Akkali1,
  8. Aruj Khurana1,
  9. Rohit Shetty1
  1. 1Narayana Nethralaya Eye Hospital, Bangalore, India
  2. 2BMS College of Engineering, Bangalore, India
  1. Corresponding to Dr Santosh G K Gadde, Department of Vitreo-retina, Narayana Nethralaya, #121, Chord Road, 1st ‘R’ Block, Rajaji Nagar, Bangalore, Karnataka 560010, India; drsantoshgk{at}


Background Macular telangiectasia type 2 (MacTel 2) is a bilateral idiopathic, rare neurodegenerative disease with alterations in the macular capillary network leading to vision loss and is the most common of three subtypes. Optical coherence tomography angiography (OCTA) is a non-invasive imaging modality which helps understand the complex pathological changes, and images the blood vessels across different layers based on their flow characteristics.

Methods A cross-sectional study was conducted on 56 eyes of Asian Indian eyes of 28 consecutive patients with MacTel 2 studied during a 3-month period in a tertiary eye care hospital of South India. Clinically diagnosed cases of MacTel 2 underwent fundus photography, spectral domain OCT and OCTA. Fluorescein angiography was performed only when clinically indicated. Mean capillary density was calculated using a MATLAB-based automated software. The images were thresholded and binarised to derive the mean value.

Results The mean age at presentation was 60±5.2, with a female preponderance of 71.42%. Vascular network on OCTA shows an increase in the intervascular spaces with progressive capillary rarefaction and abnormal capillary anastomosis. The outer retina and choroid were involved during the later stages and showed a prominent vascular network. The mean capillary density of the superficial and deep layers was 39.99% and 39.03% as against 45.18% and 44.21% in the controls, respectively. There was a statistically significant difference between the two groups (p<0.01). There is a positive and statistically significant correlation between the superficial and deep layers.

Conclusion OCTA helps understand the pathology and disease progression better in MacTel 2.

  • Imaging
  • Macula
  • Retina

Statistics from

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.


Macular telangiectasia (MacTel) is characterised by abnormalities of the capillaries around the fovea. The condition can be a result of a retinal vascular disease or systemic diseases such as diabetes or hypertension, but is most often idiopathic. MacTel was formerly called idiopathic juxtafoveolar retinal telangiectasis and was classified into three groups.1 ,2 Recent advances in imaging have allowed better characterisation of the disease to define it into two distinct forms, type 1 and type 2 MacTel.3

MacTel type 2 (MacTel 2) is a bilateral disease characterised by changes in the capillary network and neurosensory atrophy. Slit lamp biomicroscopy reveals reduced retinal transparency, crystalline deposits, ectatic capillaries and blunted venules.4 Complications of this disorder include proliferation of pigment plaques, photoreceptor loss, foveal atrophy and retinal neovascularisation that may result in visual loss.5

Stereoscopic colour and red-free fundus photography and fundus fluorescein angiography (FA) have been previously used to describe the disease. Optical coherence tomography (OCT) provides useful visualisation of the retinal and choroidal changes in patients with MacTel 2. The currently described spectral domain OCT signs (SD-OCT) of MacTel 2 include hyporeflective spaces in the inner and the outer retina (retinal cysts), plaques and breaks in the ellipsoid zone with progressive outer retinal atrophy in which the retinal layers interior to the outer nuclear layer seemingly ‘collapse’ through these breaks towards the retinal pigment epithelium (RPE).6 ,7 Progression of the disease is characterised by shrinkage of the outer retinal layers and reduction in the central foveal thickness (CFT).8 En face OCT also enables visualisation of the pigment plaques and new vessels in the various retinal layers along with the extent of associated exudation.5 ,9

FA is the gold standard to confirm the diagnosis of MacTel 2 showing telangiectatic capillaries predominantly temporal to the foveola10 in the early phase and a diffuse hyperfluorescence in the late phase. The late hyperfluorescence can be seen even in the absence of telangiectatic alterations. Gass and Oyakawa1 on studying stereoscopic angiographic images have suggested that vascular alterations mainly affect the deeper capillary plexus and that the late diffuse fluorescence on FA appears to originate from the outer retina. The finer morphological changes in the vascular network are obscured by the diffuse hyperfluorescence and FA fails to provide details about the deeper retinal layers and finer pathological changes. Besides, FA is an invasive imaging modality with documented side effects and repeating the test at every visit may not be feasible.

OCT angiography (OCTA) is a new imaging modality that allows high-resolution imaging of the retinal morphology, especially of individual vascular layers, without the use of an injectable dye. It has an A-scan rate of 70 000 scans/s, using a light source centred on 840 nm and a bandwidth of 45 nm. The tissue resolution is 5 μm axially with a 22 μm beam width. Each B-scan contains 512 A scans.11 RTVue XR 100 (Avanti, Fremont, California, USA) uses the split spectrum amplitude decorrelation algorithm to acquire information on the flow in the retinal and choroidal vasculature.12


The aim of the study was to describe the OCTA features of the different stages of MacTel 2. The capillary density of the superficial and deep layers of retina were studied and compared with controls.

Patients and methods

Study design

It is a cross-sectional, observational study.

OCTA was performed in 56 Asian Indian eyes of 28 consecutive patients with clinically confirmed MacTel 2 for a period of 3 months (February–April 2015) in a tertiary eye care hospital in South India. Ten eyes of five age-matched controls with a normal ophthalmic examination were included in the study. The study met the approval of the Institute Research Board and the Institute Ethics Committee and informed consents were obtained from all enrolled subjects. The study adhered to the tenets of the Declaration of Helsinki.

All patients underwent a complete ophthalmic examination along with fundus photography, OCT (Heidelberg engineering, Spectralis (V.6.0) GmBH) and OCTA (Optovue RTVue XR 100 Avanti, Fremont, California, USA). FA (Heidelberg engineering, Spectralis (V.6.0), GmBH) was performed only in advanced cases when clinically indicated. The Gass and Blodi classification4 was used for classifying subjects into the five clinical stages.

A single technician acquired a mydriatic OCTA with two consecutive scans of 3 by 3 mm centred on the macula with an internal fixating target. The OCTA images were independently graded and assessed by two clinicians. Segmentation of the superficial, deep, outer retina and choroidal levels was done with a default standard offset with segmentation layers based at internal limiting membrane, inner plexiform layer and RPE (figure 1). In advanced cases of MacTel 2 where the retinal architecture was altered with gross foveal thinning and extensive cystic spaces, the offset was manually adjusted to acquire the right segmentation of layers before analysis.

Figure 1

Offset and default segmentation of normal retina. ILM, internal limiting membrane; IPL, inner plexiform layer; RPE, retinal pigment epithelium.

The outer retina showed speckled pattern occasionally secondary to an artefact or noise which was minimised with the ‘remove artefacts’ option of the software and by adjusting the offset to the possible limit in case of architectural changes, based on the B scan. Other occasional artefacts included superimposed vessel imprints across different plexus. This was overcome by manual segmentation to some extent.

A MATLAB-based, semi-automated software called Ex_GUI was designed indigenously for the quantification of capillary density in the perifoveal region. A 3 mm diameter circular area centred on the fovea was considered and the capillary density of circular area was calculated by the automated software. The green channel of the RGB image was extracted, as the green channel component has the highest contrast. Image enhancement was performed on this extracted image, after the selection/demarcation of the area of the study. The contrast was increased after background subtraction to remove the artefacts. The Otsu's thresholding was used to binarise the resultant image. The final output gave a measure of the total area of study, capillary area and percentage of capillary density in pixels (figure 2).

Figure 2

Pictorial representation of Ex_GUI software.

Statistical analysis

For statistical analysis, IBM SPSS Statistics 22 was used. Since all variables studied exhibited a normal distribution, parametric tests were conducted. Post-hoc tests were applied wherever necessary. Pearson's correlation coefficients were calculated to investigate bivariate relationships and partial correlation coefficients were calculated when controlling for confounding variables. We used the 5% level of significance throughout our analysis. Independent samples t-test was performed to compare the capillary density between patients with MacTel 2 and controls. One-way analysis of variance was performed to evaluate the difference in the capillary density of the different stages of the disease.


The demographic and vision-related data of the 28 subjects recruited into the study are summarised in table 1. The age of the patients in the cohort ranged from 42 to 75 years with a mean±SD of 60±5.2.

Table 1

Demographic details

OCTA features

The OCTA features of each layer are described. Early changes begin in the superficial and deep vascular network progressing outwards with disease severity. Manual segmentation was applied when the CFT was less than 150 μ and in grossly distorted maculae.

Superficial layer

Stages 1–3 (figure 3A–C) had enlargement of the foveal avascular zone (FAZ), beginning temporally with a progressive horizontal increase in the FAZ diameter in the later stages. The perifoveal capillaries showed abnormalities in the network, the vessel size and their arrangement. There was an increase in the intervascular space with progressive capillary rarefaction. These changes were accompanied by abnormal anastomoses between the capillary networks of the superficial and deeper layers. The finer vascular anastomoses could be appreciated as discrete bunches at multiple areas, more in the temporal aspect of the macula. The abrupt ending of these vessels could be due to a gross decrease in the finer capillary network. The changes in larger blood vessels were more pronounced in the later stages with an abrupt ending and posterior dipping which could be traced into the deeper layers with adjustment of segmentation.

Figure 3

(A) Stage1: Colour fundus photo showing dull foveal reflex and optical coherence tomography (OCT) reveals intraretinal hyporeflective spaces with changes in photoreceptor layer. OCT angiography (OCTA)-macular telangiectasia type 2 (MacTel2) showing mild irregularity of the foveal avascular zone (FAZ) and minimally increased intervascular spaces more in the superficial plexus (*) and broken regular network of FAZ with broken twig like capillaries at deep plexus (arrowhead). All other layers appear normal and so are not illustrated. (B) Stage 2: Colour fundus photo showing altered fovea with slight greying of perifoveal area with OCT showing intraretinal hyporeflective spaces and internal limiting membrane drape at the fovea. On OCTA, besides stage 1 changes, there is progressive temporal enlargement and capillary rarefaction of FAZ in superficial layers (dotted circled area).Deep layer shows decrease in perifoveal capillary density but more pronounced and brighter capillary twigs, lost FAZ ring (dotted circle) and few aneurysmal dilatations (arrow). Outer retina and choroid appears uninvolved. (C) Stage 3: Colour fundus image showing vascular telangiectasia with greying of perifoveal retina. OCT shows decreased central foveal thickness, intraretinal hyporeflective spaces and photoreceptor loss. OCTA shows gross segmental loss of perifoveal capillary network, more temporally in both superficial and deep layers with increase in the FAZ and also noticeable distortion, more in the deeper plexus (dotted circled area). The right-angled vessel dipping can be appreciated and traced from superficial to deep plexus (red arrows). Outer retina and choroid were uninvolved.

Stages 4 and 5 (figures 4 and 5A, B) demonstrated a gross decrease in retinal thickness and extensive empty spaces show prominent larger blood vessels with radiating branches and widespread dark grey empty areas. The right-angled dipping of the larger blood vessels was better defined with surrounding darker areas corresponding to RPE hyperplasia in atrophic retinae.

Figure 4

Stage 4: Colour fundus photo showing pigment clumps and retinal pigment epithelium (RPE) alteration at macula. Optical coherence tomography (OCT) shows intraretinal hyper-reflectivities with backscattering and disruption of photoreceptor layer with variable retinal thinning. OCT angiography (OCTA) shows increased foveal avascular zone diameter and intervascular spaces, more horizontally with complete loss of perifoveal regular capillary meshwork in the superficial and deep layers with prominent larger vessels. A brush-like localised branching of vessels (dotted circles) with right-angled dipping can be seen around darker areas (arrows) suggestive of RPE hyperplasia. Outer retina reveals blood vessels corresponding to areas around RPE hyperplasia (arrows). Choroid reveals dark areas surrounded by gross alteration of texture surrounding it. Red arrows depict the anastomosis between the superficial and deep layers.

Figure 5

(A) Stage 5: A predominantly scarred neovascular membrane due to macular telangiectasia type 2 as seen on colour fundus photo, mid-phase frame of fluorescein angiography and optical coherence tomography scan. (B)Stage 5: Superficial and deep layers show complete distortion of foveal avascular zone, gross capillary rarefaction and dilated vessels that can be traced through different layers with abnormal anastomosis (dotted circles). Outer retina shows dense vascular network corresponding to the area of the scar (dotted circles) with an additional area of network of vessels (arrow) not corresponding to other layers suggestive of extensive areas of involvement. Choroid shows gross alterations in the texture with porous/spongy appearance, larger blood vessels with interspersed finer vessels in the area of the scar (larger dotted circle). Red arrows depict the anastomotic vessels between the superficial and deep network.

Three patients of stage 5 showed a choroidal neovascular membrane (CNVM) with scarring, significant decrease in capillaries with prominent larger blood vessels and distorted anatomical landmarks.

Deep layer

The deeper layer showed similar changes as in the superficial layer with few pathognomonic alterations. The deeper layer demonstrated changes earlier than the superficial layer in the initial stages. The normal close-knit ring of perifoveal capillaries decreased in density, but appeared more pronounced and brighter (figure 3B). There was a patchy loss of the network temporally to begin with and progressive loss of the entire ring leading to an increase in the FAZ diameter in the later stages. Brush-like branching pattern of smaller vessels secondary to capillary barring was evident (figure 4). The larger blood vessels lost the right-angled entry and branching pattern. Prominent larger vessel segments could be appreciated at multiple places with blunt ends suggestive of vessel dipping and gross capillary rarefaction (figure 5B).

Outer retina

The outer retina is a vessel-free dark area with very minimal bright speckles in normal individuals. The initial stages remain dark if the architecture is grossly maintained. A prominent vascular network with larger vessels that can be traced back to the superficial and deeper networks with a few bright and dark areas suggestive of scarring and RPE hyperplasia were seen in the later stages (Figures 4 and 5B). The vascular network of the CNVM and the scarring stage with a dramatic and exemplary network of larger and smaller blood vessels across different quadrants were seen. The vessels seen in these stages in the outer retina were not always well delineated in the superficial and deeper networks, suggesting a wider area of involvement than expected (figure 5B).

Choroid layer

The normal layer of choriocapillaris showed a diffuse granular texture of alternating dark and bright areas. The initial stages of the disease showed no changes except for some dark grey areas in cases with inner segment and outer segment (IS/OS) disruption and large areas of intraretinal cavitation. The later stages with RPE hyperplasia and CNVM scarring showed bright areas with altered choroidal architecture that was porous with a sponge-like or coral-like appearance with prominent larger vessels that seem embedded in the substance of choroid with a network of interspersed finer vessels corresponding to the area of the scar and the changes seen in the other layers (figure 5B).

Capillary density

The mean capillary density calculated by the Ex_GUI in patients with MacTel 2 was 39.99%±3.9% and 39.03%±4.54% in the superficial layer and deep layer, respectively. The corresponding mean capillary density in controls was 45.18%±0.84% and 44.21%±0.85%, respectively. There was a significant difference between the capillary density in patients with MacTel 2 and controls (p<0.01, table 2). There was a positive and statistically significant relationship (r=0.441) between the superficial capillary density and deep capillary density in patients (figure 6).

Table 2

Mean capillary density in normal individuals and patients with macular telangiectasia (MacTel)

Figure 6

The relationship between superficial capillary density and deep capillary density. Circles represent individual subjects in the study and the line represents the Pearsons correlation coefficient.


MacTel 2 is a bilateral disease with characteristic alterations in the macular capillary network with neurosensory atrophy.1 The hallmarks of the disease include hyporeflective intraretinal cavitations, foveal thinning, abnormal vascular anastomosis, RPE hyperplasia, CNVM, formation and vision-threatening progressive photoreceptor loss primarily attributed to the pathophysiological and biochemical changes involving the Müller cells.1 ,8 Presentation is after the fourth decade and the age in our cohort ranged from 42 to 75 years (60±5.2). MacTel group found the mean age of presentation to be 61±9 years.13 Other studies also showed mean age of presentation as 58±8 years.10 Spaide et al11 had an average age of 61.9 years. There was female sex predilection of 71.4% in our study which was similar to Clemons et al13 group, while Nowilaty et al14 had an equal distribution.

OCT features seen in our cohort such as IS/OS disruption, hyporeflective areas in the inner retinal layers and pigment plaques are similar to earlier studies.3 ,9 The OCTA features that we appreciated in the deep and superficial networks such as abnormal vascular anastomosis and loss of capillaries have been described on fundus FA and OCT by Yannuzzi et al.3 The abnormal vascular anastomosis described by Chin et al15 in earlier stages were seen in our study as well. OCTA observations in our study have a positive correlation with earlier histopathological studies16 that showed thickening of retinal capillaries, proliferation of the basement membrane in a multilayer configuration especially in the deep layer evident as brighter and more pronounced blood vessels in the perifoveal FAZ. Spaide et al11 showed changes in the inner and outer retinal plexus such as telangiectatic vessels with increased intervascular spaces, and the patchy loss of capillary network that began temporally and progressed with advanced stages which was seen in our cohort as well.

Imaging modalities like FA and OCT can help in gross estimation of changes in the vascular network. Limited visualisation of the deeper retinal layers on FA compounded by obscuration of finer details due to dye leakage prevents more detailed analysis. OCTA has the unique advantage of allowing visualisation of the pathology at various retinal layers in vivo,17 mimicking histopathological sections. The superficial as well as deep vascular plexus showed almost equal involvement with a slight preponderance of deep vessel involvement.3 ,16 In order to quantify our findings, capillary density at various retinal layers was calculated with indigenously developed software. There was a significant reduction in the capillary density in both the superficial and deep layers when compared with age-matched controls. In our study, there was a good correlation (r=0.441) between changes that were seen at both superficial and deep levels. Outer retina was a vessel-free area except in the later stages where there was an invasion of vessels. The retinal vascular changes begin in the superficial regions and penetrate into the deep layers progressively. A further quantification of parameters with advances in imaging software would confirm this hypothesis. The present limitations of OCTA are a reduced field of imaging and difficult quantification, which would be overcome with advances in analysis software and technology. Since MacTel is a pathology that involves the central retina, this was not a particular limitation in our study.

The OCTA uses flow dynamics,12 while the OCT works on the principle of low coherence interferometry,18 giving only the structural details which is static in nature. The en face view of the OCTA clearly depicts the vascular network, pattern of loss and sectoral involvement, thereby providing a better understanding of the pathogenesis of the disease with finer visualisation. A combination of both the modalities may be an alternative to invasive FA.


OCTA is a non-invasive, novel-imaging tool with in vivo imaging properties that are comparable with histopathological sections and helps in a better understanding of the pathogenesis, prognostication of the disease and improved standard of care. The expanded indications of OCTA will be evident after its use in other retinal disease is validated.



  • Competing interests None declared.

  • Patient consent Obtained.

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

  • Ethics approval Institutional Ethics committee approval has been received for the imaging study based on OCTA.

Linked Articles

  • At a glance
    Keith Barton James Chodosh Jost Jonas