14

Nonexudative Neovascular Age-Related Macular Degeneration

Steven T. Bailey, MD; Ching J. Chen, MD; and Yali Jia, PhD

Age-related macular degeneration (AMD) is one of the leading causes of irreversible vision loss.¹ The majority of vision loss is due to neovascular AMD, characterized by the formation of choroidal neovascularization (CNV). CNV refers to pathologic angiogenesis and vasculogenesis from the choroid that perforates Bruch’s membrane (BM) and gains access to the outer retina, between the BM and the outer plexiform layer (OPL), a region devoid of blood vessels in healthy eyes. CNV can result in exudation, hemorrhage, and fibrosis formation damaging the outer retina resulting in vision loss.²

DETECTION OF CHOROIDAL NEOVASCULARIZATION: FLUORESCEIN ANGIOGRAPHY VS OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY

In current clinical practice, the detection and diagnosis of CNV relies on fluorescein angiography (FA).³ CNV endothelium is incompetent in allowing fluorescein molecules to exit the vasculature, resulting in characteristic hyper-fluorescence patterns allowing CNV diagnosis. Therefore, by definition, CNV detected with FA is exudative. Patterns of hyper-fluorescence changing over time are used to classify CNV as classic, occult, or both. Classic CNV has well-defined borders in earlier frames of the FA and late hyper-fluorescent leakage. Occult CNV often appears ill defined or is difficult to detect with FA because the retinal pigment epithelium (RPE) blocks the transmission of fluorescence. Indocyanine green angiography (ICGA) improves the detection of occult CNV,⁴ but is used less frequently clinically. Drawbacks to dye-based angiography include long acquisition time ranging from 5 minutes to 20 minutes and possible allergic reaction, including anaphylaxis.⁵

OCTA is a noninvasive approach to visualize blood vessels of the retina and choroid and does not rely on dye-leaking patterns for CNV detection.⁶ Instead, the 3-dimensional (3D) nature of OCTA allows CNV to be detected as pathologic blood flow in the outer retinal/RPE slab, between the outer boundary of the OPL and BM. Clinicians can often identify CNV with en face imaging because CNV has unique branching patterns compared to retinal and choroidal circulations. Combined cross-sectional structural OCT with OCT angiograms localize the depth of the CNV as either below the RPE (type 1 CNV) or above the RPE (type 2 CNV). Several studies have demonstrated the role of detecting CNV with OCTA in AMD7–14 and other diseases, including central serous chorioretinopathy^15–17 and inflammatory disease.¹⁸ In addition to detection of CNV, novel quantification metrics are possible.^7,11,19 A CNV area is generated with OCTA by summing pixels with blood flow detected in the outer retinal/RPE slab and can be used as a metric to monitor CNV growth or response to treatment.²⁰

NONEXUDATIVE NEOVASCULAR AGE-RELATED MACULAR DEGENERATION

Nonexudative CNV refers to CNV visible with OCTA in the absence of fluid or hemorrhage on clinical examination and the absence of leakage on FA.²⁰ With structural OCT, there is no retinal thickening and no signs of intraretinal fluid (IRF) or subretinal fluid (SRF). Structural OCT will reveal a separation between BM and RPE accounting for the space for the CNV to occupy. OCTA can detect moving red blood cells in this location, confirming the presence of CNV.

There are several possible variants of nonexudative CNV that can be segregated as either treatment naïve, or previously treated CNV. Treatment-naïve nonexudative CNV occurs when pathologic vessels from the choroid grows through the BM into the sub-RPE space; however, these vessels have either low flow or adequate capillary endothelium to prevent fluid from exuding into the retinal tissue, and visual function is not affected. Previously treated nonexudative CNV refers to CNV that has responded to anti-vascular endothelial growth factor (VEGF) therapy. Treatment-responsive CNV may include cases that require ongoing intermittent anti-VEGF treatments to prevent exudation and CNV growth. CNV in remission refers to treated CNV that has had no recurrence of exudation for greater than 12 months without treatment.

Clinical Examples of Nonexudative Choroidal Neovascularization

Five different examples of nonexudative CNV are presented in this chapter. The OCTA methodology for each case was the same. OCT angiograms were acquired using the commercially available spectral domain (SD) OCT (RTVue XR, Optovue Inc) along with the split-spectrum amplitude-decorrelation angiography algorithm. The dataset was then exported for custom processing at the Casey OCT Angiography Reading Center (Portland, OR). The 3D OCT angiogram was segmented into individual vascular layers that were color-coded: inner retinal slab (purple), choroidal slab (red), and outer retinal/RPE slab representing CNV (yellow).¹¹ Semiautomated advanced image processing techniques were used for CNV detection and quantitative analysis.^21–23 CNV area represents the total area of vessels with blood flow within a CNV network.

Treatment-Naïve Nonexudative Choroidal Neovascularization

Case 1: Treatment-Naïve Nonexudative Choroidal Neovascularization Adjacent to Geographic Atrophy

An 88-year-old female was diagnosed with exudative neovascular AMD in the right eye and presumed non-neovascular AMD in the left eye. Visual acuity measured 20/125 in the right eye and 20/25 in the left eye. Evaluation of the left eye revealed geographic atrophy (GA) in the superior macula and no clinical signs of fluid or hemorrhage on examination (Figure 14-1A). FA revealed a well-demarcated pattern of staining in the superior macula consistent with GA along with staining of several drusen. No late fluorescein leakage was present (Figures 14-1B and 14-1C) and findings were non-diagnostic for presence of CNV. Structural OCT of the nasal macula showed a thin separation between BM and RPE; however, no SRF/IRF was present, and therefore no clinical reason to be suspicious for CNV. En face OCT angiograms revealed a network of vessels with anastomotic loops in the nasal macula, adjacent to the GA region. These vessels were highlighted with yellow coloring demonstrating both the location and depth of the CNV (Figure 14-1D). Cross-sectional OCT and color-coded OCT angiograms demonstrate flow between BM and the RPE consistent with type 1 CNV (Figure 14-1E). Given the absence of symptoms and lack of exudation associated with this lesion, treatment was not recommended.

Case 2: Treatment-Naïve Nonexudative Choroidal Neovascularization with Dense Microvascular Network

An 88-year-old female with vision loss in the right eye due to retinal detachment was diagnosed with presumed non-neovascular AMD in both eyes. Visual acuity measured counting fingers in the right eye and 20/25 in the left eye. Fundus examination revealed drusen with mild pigmentary changes (Figure 14-2A). FA revealed several punctate areas of hyper-fluorescence stain in the later frames without definitive leakage, findings non-diagnostic for the presence of CNV (Figures 14-2B and 14-2C). Structural OCT revealed a pigment epithelial detachment (PED) present at the fovea and inferior macula. No SRF/IRF or retinal thickening was associated with the PED. With en face OCTA (Figure 14-2D), a microvascular network is visible underneath and inferior to the fovea. Isolated outer retinal/RPE slab (Figure 14-2E) with projection artifact removal allows a clean view of CNV, which appears as a dense collection of fine anastomosing vessels, an appearance with some similarities to the choriocapillaris. Cross-sectional OCT and OCT angiogram (Figure 14-2F) detected flow within the PED consistent with type 1. Given the absence of symptoms and lack of exudation, treatment was not recommended.

Case 3: Treatment-Naïve Nonexudative Choroidal Neovascularization with Longitudinal Follow-up

A 68-year-old male presented with new-onset exudative neovascular AMD in the right eye and presumed non-neovascular AMD in the left eye. Visual acuity measured 20/50 in the right eye and 20/32 in the left eye. An occult CNV was present on FA along with SRF on structural OCT. FA of the left eye revealed staining of the RPE and drusen without leakage, suggesting no CNV (Figures 14-3A to 14-3C). Structural OCT revealed the RPE was slightly elevated above BM inferior to the fovea; however, no fluid was present. OCT angiograms revealed a vascular network just inferior to the fovea in the outer retinal/RPE slab. CNV area on initial visit measured 0.20 mm². Structural OCT super-imposed with cross-sectional OCT angiogram (Figure 14-3D) demonstrated CNV as type 1.

The right eye was treated with anti-VEGF treatment and the left eye was not treated given the absence of symptoms associated with the nonexudative CNV. The patient was followed for 15 months with frequent visits, about every 4 to 8 weeks. During this time, the patient never developed visual symptoms in the left eye. Structural OCT at each visit revealed no evidence of fluid or retinal thickening. Serial OCT angiograms of the inner retina and outer retina/RPE slab demonstrated slow growth of CNV (Figures 14-4A to 14-4C), with the CNV area increasing from baseline measurement of 0.20 mm² to 0.26 mm². Segmented OCT angiogram of the outer retinal/RPE slab after projection artifact removal provides a clear view of CNV (Figures 14-4D to 14-4F), revealing an increasing number of vascular loops and anastomoses that develop over time. FA was repeated 5 and 15 months after presentation revealing no evidence of leakage to suggest the presence of CNV in the left eye (Figures 14-5A and 14-5B). ICGA at 15 months revealed several discrete multifocal areas of late staining (Figures 14-5C and 14-5D). These areas could represent small plaques consistent with sub-RPE CNV that partially correlate to CNV with OCTA (Figures 14-5E and 14-5F). Currently, this eye remains untreated.

Stay updated, free articles. Join our Telegram channel

Join