Multimodality imaging in macular telangiectasia 2: A clue to its pathogenesis.

Macular telangiectasia type 2 also known as idiopathic perifoveal telangiectasia and juxtafoveolar retinal telangiectasis type 2A is an acquired bilateral neurodegenerative macular disease that manifests itself during the fifth or sixth decades of life. It is characterized by minimal dilatation of the parafoveal capillaries with graying of the retinal area involved, a lack of lipid exudation, right-angled retinal venules, refractile deposits in the superficial retina, hyperplasia of the retinal pigment epithelium, foveal atrophy, and subretinal neovascularization (SRNV). Our understanding of the disease has paralleled advances in multimodality imaging of the fundus. Optical coherence tomography (OCT) images typically demonstrate the presence of intraretinal hyporeflective spaces that are usually not related to retinal thickening or fluorescein leakage. The typical fluorescein angiographic (FA) finding is a deep intraretinal hyperfluorescent staining in the temporal parafoveal area. With time, the staining may involve the whole parafoveal area but does not extend to the center of the fovea. Long-term prognosis for central vision is poor, because of the development of SRNV or macular atrophy. Its pathogenesis remains unclear but multimodality imaging with FA, spectral domain OCT, adaptive optics, confocal blue reflectance and short wave fundus autofluorescence implicate Müller cells and macular pigment. Currently, there is no known treatment for this condition.

Many ocular and systemic conditions may manifest retinal telangiectasis or abnormal dilation of the retinal capillary network. In 1982, Gass and Oyakawa[1] were the first to identify patients with retinal telangiectasis limited to the parafoveal area with no apparent specific cause. They named this condition idiopathic juxtafoveolar retinal telangiectasis (IJRT) and classified it into four groups. Over the years, others have referred to this condition as idiopathic parafoveal retinal telangiectasis.[1] In 1993, Gass and Blodi[2] further modified this classification by dividing the eyes into three groups and each group was further subdivided into two other sub-groups. In an attempt to simplify the Gass-Blodi classification, Yannuzzi et al.[3] divided IJRT into two broad groups: Aneurysmal telangiectasia or idiopathic macular telangiectasia type 1 (MacTel 1) and perifoveal telangiectasis, also known as idiopathic (MacTel 2). Eyes with MacTel 2 were further subdivided into the nonproliferative stage characterized by telangiectasia and foveal atrophy; and the proliferative stage characterized by the presence of subretinal neovascularization (SRNV).[3]

Our understanding of MacTel 2 has paralleled advances in multimodality imaging of the ocular fundus. For many years, MacTel 2 was considered as primary retinovascular disease based on the fluorescein angiographic (FA) findings. Multimodality imaging has provided new insights into the pathogenesis of this condition.

Clinical Findings

Most patients complain of nonspecific symptoms such as mild blurring of vision, positive scotoma, difficulty in reading and metamorphopsia.[4567] Initially, the disease is characterized by relatively good visual acuities of ≥ 20/30.[89] The earliest ophthalmoscopic changes seen in MacTel 2 are rather subtle and may be missed easily. The first sign is a mild grayish discoloration of the retina with loss of retinal transparency temporal to the fovea. At this point, telangiectatic vessels will be absent or barely evident on clinical examination. FA is often necessary to demonstrate the abnormal juxtafoveolar capillary network.[3810] With disease progression, this grayish discoloration surrounds the parafoveal retina in an oval configuration.[2310] In addition, slightly dilated and blunted retinal venules that extend at right angles appear in the temporal parafoveal retina. Multiple, crystalline, golden, tiny, refractile deposits near the inner retinal surface are a common finding occurring in up to 45% of eyes.[121011] These deposits are usually found near the telangiectasis and do not seem to correlate with the severity of the disease.[11] Intraretinal round yellow spot lesions measuring between 100 μm and 300 μm in diameter similar to those seen in the adult form of vitelliform foveomacular dystrophy or Best's disease occur in up to 5% of cases of MacTel 2.[2121314] Stellate foci of intraretinal pigmented black plaques composed of hyperplastic retinal pigment epithelium (RPE) cells may develop along the right angled vessels. A lamellar macular hole may develop as the result of focal atrophy of the foveolar retina. These lamellar holes are characterized by a distinct, often circular margin and central retinal thinning that does not extend beyond the edges of the capillary free zone.[1231516] Full-thickness macular holes have also been reported in MacTel 2.[171819202122]

Typically, there is no lipid exudation or hemorrhages associated with MacTel 2 unless SRNV is present.[29] SRNV usually develops in the vicinity of the intraretinal pigment plaques. Once present, SRNV is characterized by a rapid loss of vision, subretinal hemorrhage, cystoid macular edema, lipid hard exudates, disciform scarring, and retinochoroidal anastomosis.[313] Unlike choroidal neovascularization in age-related macular degeneration (AMD), SRNV in MacTel 2 is not usually accompanied by a RPE detachment.[3] Furthermore, the size of SRNV on MacTel 2 is small in comparison to AMD.[2]

Fluorescein Angiography

Dilated ectatic perifoveal capillaries that leak dye in the parafoveal temporal areas are the typical FA findings in MacTel 2. In the early stages of the disease, the early phases of the angiogram are characterized by minimal or no evidence of capillary dilatation. The late phases are characterized by mild staining of the temporal parafoveal retina. This staining spares the foveal center. As the disease progresses, the capillary dilatation and permeability changes in the outer retina extend beyond the temporal parafoveal area to completely surround the fovea. This is manifested as a late oval-shaped parafoveal hyperfluorescence [Fig, 1].[2] The fluorescein leakage is not related to cystic spaces.[3] FA may also demonstrate dilated right-angle vessels. Intraretinal and/or subretinal anastomosis may arise from these vessels.[3] The foveal avascular zone in eyes with MacTel 2 is significantly reduced in size when compared to normal eyes.[23] The stellate hyperpigmented plaques typically block fluorescence. SRNV, when present, usually develops in the vicinity of the stellate hyperpigmented plaques.[2813] The SRNV appears to originate from the deep retinal circulation and is characterized by early and late fluorescein leakage.[324]

Figure 1

Late frame from a fluorescein angiogram of an eye with macular telangiectasia type 2 showing the typical late leakage that does not involve the foveal center

Indocyanine Green Angiography

Indocyanine green angiography (ICG-A) is the imaging modality of choice of the choroidal circulation. There are no MacTel 2 specific findings on ICG-A.[25] This should not come as a surprise since MacTel 2 spares the choroid and choriocapillaris.

Optical Coherence Tomography

Optical coherence tomography (OCT) has played a key role in deepening our understanding of MacTel 2 by demonstrating the structural changes that occur in this disease. Unlike other retinal vascular disorders such as diabetic macular edema and branch retinal vein occlusion, OCT has shown that angiographic areas of leakage in MacTel 2 do not correlate with the presence of cystoid macular edema[162627] or retinal thickening [Fig. 2].[16272829] In fact, foveal thickness is decreased in most patients with MacTel 2.[7262728303132]

Figure 2

Spectral domain optical coherence tomography showing disruption of the ellipsoid line

In addition, intraretinal hyporeflective spaces that are usually not related to retinal thickening or FA leakage are commonly seen.[162728293031333435] The depth, size and location of these spaces influence macular function more than their mere presence.[7] Several different names for these lesions, including internal limiting membrane drape, cyst, cystoid or pseudo-cystoid space, have been used in the literature.[1628]

Several investigators have described spectral-domain (SD)-OCT abnormalities of the outer retina.[1634363738] All of these descriptions probably refer to the same pathologic process but at different stages of the disease.[36] Round, oval or comma-shaped hyper reflective spots in the outer parafoveolar layers have been described to occur prior to any FA changes.[3637] Alterations of the outer plexiform layer that have been described as “wrinkling” toward the outer retina have also been reported.[34] Thinning, disruption or loss of the photoreceptor layer particularly on the temporal side of the fovea and extending to the whole fovea in advanced cases are common SD-OCT findings [Fig. 2].[1627313538]

More recently, Chhablani et al.[39] reported that in eyes with MacTel 2 the ganglion cell-inner plexiform layers are thinned suggesting a neurodegenerative process.

Confocal Blue Reflectance

Confocal blue reflectance (CBR) imaging is a fast, safe, noninvasive imaging modality that captures the fundus reflectance after illuminating it with a confocal blue light of 488 nm emitted by a scanning laser ophthalmoscope (SLO). Unlike FA, there are no intervening barrier filters.[22] In eyes with MacTel 2, CBR imaging demonstrates an increased reflectance in an oval parafoveal pattern that is slightly larger than the area of late phase FA hyperfluorescence [Fig. 3].[223440]

Figure 3

Increased confocal blue reflectance

Fundus Autofluorescence

Short wavelength fundus autofluorescence (SW-FAF) images of the posterior pole of the eye are obtained in a similar fashion to FA. The only difference is that there is no fluorescein injection. Briefly, the fundus is illuminated with confocal blue light of 488 nm emitted by SLO. The images are captured after the reflected light is filtered through a barrier filter of 520 nm. SW-FAF of the normal eye shows a central dark area caused by the light absorption by both RPE melanin and macular pigment.[41] Lutein and zeaxanthin are the two carotenoids that constitute the macular pigment. They accumulate primarily in the macular photoreceptor axons. In the parafoveal region, the inner and outer plexiform layers manifest the greatest concentration of macular pigment whereas in the fovea the greatest concentration would be in Henle's layer.[42]

One of the earliest changes reported in MacTel 2 is an increased SW-FAF signal in the foveal region [Fig. 4].[43] Initially, eyes exhibit a triangular segment of reduced macular pigment in the temporal fovea and central accumulation of macular pigment.[44] Then, there is further expansion of the triangular segment and disappearance of the central accumulation.[45]

Figure 4

Increased short wave fundus autofluorescence

Adaptive Optics

As a wavefront of light passes through the pupil, ocular aberrations are normally induced. The quality of the image formed on the retina is diminished by these aberrations. Adaptive optics (AO) corrects these ocular aberrations and improves transverse resolution in retinal imaging.[46] AO allows high-resolution photoreceptor imaging.[464748]

Adaptive optics documents disruption of the normal cone mosaic pattern in eyes with MacTel 2. Instead of the normal cone mosaic, eyes with MacTel 2 exhibit a lower cone density and ring-like or patchy dark areas on AO imaging. These abnormalities were even present in areas with normal retinal vasculature that strongly suggests that the neural degeneration precedes the vascular changes.[49] The dark areas on AO imaging corresponded to the late FA leakage. In some eyes, these AO dark areas occurred in areas without FA leakage. Disruption of the ellipsoid line on SD-OCT was also associated with the dark regions on AO imaging.[49]

Histopathology

There are currently only two clinicopathologic studies of confirmed MacTel 2 cases reported in the literature.[5051] Luteal pigment was characteristically absent in these eyes. In addition, immunohistochemical analysis demonstrated the loss of perifoveal Müller cells. Interestingly, there was a topographical correlation between areas of macular pigment absence and areas of Müller cell depletion.[50] Loss of the ellipsoid line seen in SD-OCT was correlated with rod depletion.[51] Macular pigment loss occurs prior to the ellipsoid abnormalities seen on SD-OCT.[51]

Animal Models

A transgenic mouse model with conditional ablation of Müller cells has been developed.[52] These animals exhibit photoreceptor apoptosis, retinal telangiectasis, breakdown of the blood retinal barrier and intraretinal neovascularization.

Pathogenesis

All the available evidences point to the Müller cell as a central player in MacTel 2. The findings on AO imaging clearly demonstrates that neural degeneration precedes retinal vascular involvement.[49] The increased SW-AF signal in the fovea also precedes the FA changes. Histopathological findings show that the RPE is healthy in MacTel 2.[5354] Thus, the increased SW-AF signal is most likely due to the depletion of luteal pigment rather than an increased lipofuscin accumulation in the RPE. Müller cells serve as a retinal reservoir for xanthophyll. Therefore any pathological process that involves Müller cells will affect the luteal pigment.[55] The increased CBR signal also implicates Müller cells. The increased CBR may be explained by four different mechanisms. First, since the absorption maximum of macular pigment is in the range of blue light at approximately 460 nm, decreased absorption or increased reflection of the blue light may be secondary to a decrease in macular pigment in the parafoveal area.[2256] Second, Müller cells span the entire retinal thickness and may serve as optical fibers that allow transmission with a minimal reflection of light across the retina. Any pathological involvement of Müller cells would interrupt this mechanism and decrease transmission and increase reflectance.[57] Third, disruption of the normal architecture of the neurosensory retina coupled with the presence of edema may increase retinal reflectance. This does not seem to be the case in MacTel 2 since increased CBR may present without any corresponding OCT findings.[34] Finally, defective storage or metabolism of macular pigment may cause the formation of highly reflective crystalline deposits in the neurosensory layers. This may lead to an increased CBR.[22] So far, there have been no reports of such crystalline deposition but it may well be that these deposits are below the current detection level threshold of 7 μm of axial resolution.[34]

Since Müller cells play an important role in the structural integrity of the fovea, the SD-OCT findings of hyporeflective spaces may well be cavitary spaces representing Müller cell loss rather than fluid-filled cystic spaces.[3] In addition, Müller cells help provide nutrition to the surrounding retinal neurons and also play a role in inducing and maintaining the integrity of the blood retinal barrier.[55] Their processes are intimately related to the retinal blood vessels in the outer plexus. The transgenic mouse model with conditional ablation of Müller cells demonstrates all the clinical findings seen in MacTel 2.[52]

Unresolved Issues

It remains unclear why some eyes develop SRNV and others do not.

It remains unclear as to why the temporal parafoveal region is affected first.

There are currently no treatment options for this disease with the exception of anti-vascular endothelial growth factor agents for eyes with SRNV.

Take Home Message

MacTel 2 is a primary neurodegenerative macular disease with secondary vascular involvement.

Future Directions

Development of the therapeutic agents that rescue Müller cells to test this hypothesis.

Observations using new imaging tools such as OCT angiography with its ability to image the deeper retinal capillary plexus will undoubtedly provide new insights in the progression of the disease.

Salient Features

One of the earliest changes reported in MacTel 2 is an increased SW-FAF signal in the foveal region

CBR imaging demonstrates an increased reflectance in an oval parafoveal pattern that is slightly larger than the area of late phase FA hyperfluorescence

AO imaging clearly demonstrates that neural degeneration precedes retinal vascular involvement

Clinicopathological correlation demonstrates that there is loss of perifoveal Müller cells that topographically correlate with areas of luteal pigment loss

A transgenic mouse model with conditional ablation of Müller cells exhibits many of the clinical findings seen in MacTel 2.

Literature Search

A search on MEDLINE and old MEDLINE using multiple search words including retinal telangiectasis, juxtafoveolar telangiectasis, parafoveal telangiectasis, and macular telangiectasia was peformed on August 31, 2014. Articles and book chapters cited in the reference lists of articles obtained by this method were reviewed and included when considered appropriate. Relevant articles published in the English, French and Spanish languages in peer-reviewed journals were also included.