Purpose
To determine the proportion of shallow irregular pigment epithelial detachments in eyes with pachychoroid features that harbor neovascular tissue and to study the morphology of this tissue with optical coherence tomography (OCT) angiography.
Design
Prospective consecutive cohort study.
Methods
Patients with pachychoroid spectrum diagnoses and shallow irregular pigment epithelial detachment in at least 1 eye (study eye) were included. Charts and multimodal imaging were reviewed to determine a dye angiography detection rate for type 1 neovascularization in study eyes. All patients then underwent OCT angiography prospectively, followed by masked segmentation and grading.
Results
Twenty-two eyes of 16 patients were included. Mean age was 71 (range 57–95) years. Mean subfoveal choroidal thickness was 381 μm (standard deviation: 141 μm). Four out of 22 study eyes (18%) exhibited polypoidal lesions. Dye angiography demonstrated specific features of neovascularization in 5 out of 17 eyes (29%) with suspected nonpolypoidal pachychoroid neovasculopathy. With OCT angiography, type 1 neovascular tissue was visualized in 21 out of 22 study eyes (95%).
Conclusions
Our data suggest that, in eyes with pachychoroid features, the finding of a shallow irregular pigment epithelial detachment on OCT has greater diagnostic value for type 1 neovascularization than previously thought and that dye angiography may underestimate the prevalence of neovascularization compared to OCT angiography.
Retinal pigment epithelial detachments (PEDs) are seen most commonly in age-related macular degeneration. They also occur in 70%–100% of eyes with central serous chorioretinopathy with various morphologies and reflectivity characteristics, some of which may resemble those of neovascular age-related macular degeneration.
Type 1 neovascularization is defined by the presence of a vascularized PED. Diagnosis of active neovascularization relies on cross-sectional optical coherence tomography (OCT) to image the PED and fluorescein angiography to demonstrate “poorly defined” stippled hyperfluorescence with late leakage and/or staining at the site of the PED. Type 1 neovascularization may also occur in a more benign or “quiescent” form in which fluid and leakage are not detectable by OCT or fluorescein angiography, but a shallow irregular PED is seen with a corresponding plaque in the late phase of an indocyanine green angiogram. Shallow irregular PEDs were previously described in the polypoidal literature in terms of a “double-layer sign” resolved by OCT. These PEDs were shown by Sato and associates to contain the type 1 neovascular tissue representing branching vascular networks that feed polypoidal lesions.
The term “pachychoroid” describes a set of choroidal characteristics shared by a group of related diseases. These features, which include focal or diffuse increase in choroidal thickness, choroidal hyperpermeability, and dilated choroidal vessels, were first described using indocyanine green angiography and enhanced depth imaging OCT in patients with chronic central serous chorioretinopathy. Recently, en face imaging with swept-source OCT has revealed new findings that refine the definition of the pachychoroid phenotype to emphasize the morphologic characteristics of pathologically dilated choroidal vessels (“pachyvessels”) over absolute choroidal thickness.
Pachychoroid neovasculopathy refers to type 1 neovascularization in patients who do not have drusen or other risk factors for neovascularization but who exhibit pachychoroid features. Although type 1 neovascularization is a well-recognized complication of chronic central serous chorioretinopathy, it can also arise in pachychoroid eyes that have not yet manifested frank subretinal fluid. Pigment epithelial detachments in these eyes often have a shallow irregular morphology with moderate internal reflectivity on OCT that closely resembles the “double-layer sign.” However, in the setting of pachychoroid, choroidal hyperpermeability and chronic alterations of the retinal pigment epithelium (RPE) and outer retina can compromise the diagnostic specificity of dye-based angiography for type 1 neovascularization, leaving the clinician with some uncertainty as to the vascularity of these PEDs.
Optical coherence tomographic angiography images microvascular circulation in vivo without the need for intravenous dye or contrast injection. Spaide and Kuehlewein and associates have described the distinctive OCT angiography appearance of type 1 neovessels in age-related macular degeneration treated with anti–vascular endothelial growth factor (anti-VEGF) agents and have presented explanations to account for their large caliber and de-arborized appearance.
The purpose of this paper is to determine the proportion of shallow irregular PEDs in pachychoroid eyes that contain neovascular tissue and to study their morphology with OCT angiography.
Methods
This prospective cohort study was approved by the Western Institutional Review Board (Olympia, Washington, USA). It complies with the Health Insurance Portability and Accountability Act of 1996 and follows the tenets of the Declaration of Helsinki.
Patients were recruited consecutively from the practice of a single retina specialist (K.B.F.) if they manifested a shallow irregular PED in the macula of at least 1 eye in the presence of pachychoroid characteristics, defined as any of the following: (1) choroidal thickness in excess of 270 μm, (2) the presence of pachyvessels, (3) history of central serous chorioretinopathy. In order to maintain a consistent phenotype, patients were excluded if they exhibited macular drusen, if they had any history of inflammatory eye disease, if they were highly myopic, or if image quality was too poor for grading.
Clinical charts and previous multimodal imaging data were reviewed. Patients had previously undergone fundus photography, fluorescein and/or indocyanine green angiography, fundus autofluorescence, and OCT as clinically indicated for their ongoing care. Some patients had also had swept-source OCT. Instruments used were the TRC-50IX flood-illuminated fundus camera (Topcon Corp, Tokyo, Japan), the OCT Stratus 3000 (Carl Zeiss Meditec Inc, Dublin, California, USA), the HRA-OCT scanning laser ophthalmoscope with OCT (Heidelberg Engineering, Heidelberg, Germany), and the DRI OCT-1 (“Atlantis,” Topcon Medical Inc, Oakland, New Jersey, USA). Choroidal thickness measurements were taken using the software caliper tool of the Heidelberg Eye Examination software on enhanced depth scans.
All patients underwent OCT angiography (RTVue XR “Avanti,” Optovue, Fremont, California, USA). The RTVue XR is a spectral-domain OCT device that performs 70,000 A-scans per second and employs a split-spectrum amplitude decorrelation algorithm to produce depth-resolved 3-dimensional microvascular volume maps of the posterior pole. The OCT angiography volume scans were segmented manually using the viewing software accompanying the RTVue XR and were examined by a masked grader (C.B.) for the presence or absence of type 1 neovascular tissue with tangled vascular morphology, as previously described. The en face area of neovascular complexes was determined by manual border tracing using analysis tools in the Fiji distribution (“Fiji is just ImageJ,” http://fiji.sc ).
Since dye angiography had been performed only when clinically indicated, the time interval between dye angiography and OCT angiography was not controlled. Where dye angiograms were available for multiple historic time points for any given patient, the angiogram contemporary with the onset of the shallow irregular PED was graded. Owing to the relatively recent availability of OCT angiography in our practice (approximately 1 year), it was not possible to obtain OCT angiographic data contemporary with the onset of shallow irregular PED for several patients in this cohort.
Results
Baseline Demographics
A total of 22 eyes of 16 patients met the selection criteria. The mean age of the patients was 71 years (range 54–95 years). Eight patients were male and 8 were female. Shallow irregular PEDs were noted in all eyes and were bilateral in 6 out of 16 patients. The Table includes a list of patients with demographic data and visual acuities at final follow-up.
| Patient # | Age | Sex | Eye | Eye # | Snellen VA | Polypoidal Lesions | Fluorescein Angiography | Indocyanine Green Angiography | Neovascular Detection by Dye Angiography | Dye-Angiography → OCT Angiography Interval (Months) | Neovascularization on OCT Angiography | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Scan Area (mm 2 ) | Morphology | Trunk Diameter (μm) | Neovascular Complex Area (mm 2 ) | |||||||||||
| 1 | 63 | F | OD | 1 | 20/25 | Nonspecific | Nonspecific | Negative | 1 | 9 | Tangle | 84 | 0.594 | |
| OS | 2 | 20/20 | Nonspecific | Not available | Negative | 1 | 9 | Loop | Not identified | 0.087 | ||||
| 2 | 86 | M | OD | 3 | 20/70 | Nonspecific | Not available | Negative | 84 | 9 | Tangle | 86 | 2.968 | |
| OS | 4 | 20/40 | Not available | Not available | Not available | 9 | Placoid | Not identified | 0.303 | |||||
| 3 | 70 | F | OD | 5 | 20/70 | Present | Polypoidal + masking of choroidal neovascularization | Branching vascular network + polypoidal | Positive | 108 | 9 + 36 (36 mm 2 montage) | Tangle | Not identified | >4.51 |
| 4 | 63 | F | OD | 6 | 20/100 | Present | Nonspecific | Polypoidal | Positive | 4 | 36 + 36 (58 mm 2 montage) | Tangle | 146 | >20.0 |
| OS | 7 | 20/25 | Present | Nonspecific | Choroidal neovascularization + polypoidal | Positive | 4 | 9 | Tangle | 129 | >4.53 | |||
| 5 | 60 | M | OD | 8 | 20/60 | Normal | Nonspecific | Negative | 26 | 9 | Tangle | Not identified | 0.729 | |
| OS | 9 | 20/60 | Choroidal neovascularization | Plaque | Positive | 26 | 9 | Tangle | 46 | >3.728 | ||||
| 6 | 70 | F | OS | 10 | 20/25 | Not available | Plaque | Positive | 9 | 9 | Tangle | 57 | 0.669 | |
| 7 | 76 | M | OS | 11 | 20/40 | Atrophy | Not available | Negative | 39 | 9 | Tangle | 69 | 3.719 | |
| 8 | 67 | M | OD | 12 | 20/30 | Choroidal neovascularization | Plaque | Positive | 60 | 9 | Tangle | 60 | 1.415 | |
| 9 | 58 | F | OS | 13 | 20/25 | Present | Not available | Choroidal neovascularization + polypoidal | Positive | 45 | 9 | Tangle | 36 | 3.480 |
| 10 | 57 | M | OD | 14 | 20/20 | Nonspecific | Nonspecific | Negative | 34 | 9 | Tangle | 117 | >6.44 | |
| OS | 15 | 20/30 | Nonspecific | Nonspecific | Negative | 49 | 9 | Tangle | Not identified | >6.85 | ||||
| 11 | 70 | F | OS | 16 | 20/50 | Choroidal neovascularization | Plaque | Positive | 19 | 9 | Tangle | 48 | 3.76 | |
| 12 | 54 | F | OS | 17 | 20/25 | Choroidal neovascularization | Nonspecific | Positive | 19 | 9 | Tangle | 87 | 0.512 | |
| 13 | 95 | M | OD | 18 | 20/70 | Nonspecific | Not available | Negative | 79 | 9 | Tangle | Not identified | 0.463 | |
| OS | 19 | 20/30 | Nonspecific | Not available | Negative | 79 | 9 | Tangle | 49 | 0.544 | ||||
| 14 | 74 | M | OD | 20 | 20/40 | Nonspecific | Not available | Negative | 19 | 9 | Not detected | Not identified | ||
| 15 | 70 | F | OS | 21 | 20/30 | Nonspecific | Nonspecific | Negative | 27 | 9 | Tangle | 50 | 0.777 | |
| 16 | 82 | M | OS | 22 | 20/50 | Atrophy | Not available | Negative | 20 | 9 | Tangle | Not identified | 1.479 | |
Nineteen study eyes had a history of prior treatment. Eighteen eyes had received intravitreal anti-VEGF treatment (mean 31.3 injections per eye; standard deviation [SD]: 22.7) and 6 eyes had received verteporfin photodynamic therapy (mean 2.6 treatments per eye; SD: 1.5).
Choroidal Features
Mean subfoveal choroidal thickness for study eyes was 381 μm (SD: 141 μm). Seven of the study eyes had choroidal thicknesses less than 270 μm (mean: 212 μm; range: 176–248 μm; SD: 28.8 μm), but each of these qualified as pachychoroid (as opposed to age-related macular degeneration) on the basis of one or more previously described criteria : (1) an extrafoveal focus of maximal choroidal thickness exceeding subfoveal choroidal thickness by at least 60 μm (2 standard deviations) (2 eyes); (2) pachyvessels with distinctive en face OCT morphology (7 eyes). Four of these eyes had previously had photodynamic therapy.
Dye Angiography
Fluorescein angiography was available for 19 study eyes and revealed features of type 1 neovascularization in 4 eyes. In 14 study eyes the pattern of hyperfluorescence could not be distinguished from that seen in chronic central serous chorioretinopathy, and type 1 neovascularization could not be diagnosed specifically. Fluorescein angiography was normal in 1 study eye. Indocyanine green angiography was available for 14 study eyes and revealed hyperfluorescent plaques in 4 eyes and polypoidal lesions with branching vascular networks in 4 eyes. In the 4 eyes with polypoidal lesions, grading for the presence or absence of type 1 plaques on indocyanine green angiography was difficult to perform objectively because the grader could not be masked to the presence of polypoidal lesions. For study eyes with nonpolypoidal pachychoroid neovasculopathy that had at least one form of dye-based angiography, the combined detection rate for type 1 neovascularization was 5 out of 17 eyes (29%). Dye angiographic findings are listed in the Table .
The mean interval between dye angiography and OCT angiography for nonpolypoidal study eyes was 36 months (SD: 30.5 months) and any further analysis to compare these modalities directly on the basis of data in this cohort was therefore felt to be inappropriate.
Optical Coherence Tomography Angiography
Type 1 neovascular networks were detected in 21 out of 22 study eyes (95%) by OCT angiography. In 18 of these eyes, en face segments taken through the PEDs showed tangled networks of large-caliber “trunk vessels” with identifiable feeder vessels (eg, Figure 1 , bottom row, first image); in 2 eyes, the type 1 networks were small and localized (eg, Figure 1 , bottom row, third image); in 1 eye, the flow signature of the type 1 lesion had a more consolidated or placoid morphology ( Figure 2 , second row, fourth image). In the 4 eyes with polypoidal lesions, the polypoidal lesions themselves did not exhibit a detectable flow signature. One of the study eyes did not exhibit neovascularization on OCT angiography. (One eye of 1 patient had an ungradeable OCT angiogram owing to poor fixation and motion artifact and is neither shown nor represented in any figures or analyses.)
In each of the 4 eyes that manifested polypoidal lesions, the appearance of the associated type 1 neovascular network on spectral-domain OCT closely resembled that of branching vascular networks of polypoidal choroidal vasculopathy as described by Sato and associates. Type 2 and type 3 neovascularization were not detected in any of the study eyes by any of the imaging modalities employed.
Representative Case Descriptions
In each of the following cases, en face OCT angiography through the shallow irregular PED showed a flow signature compatible with type 1 neovascularization.
Patient 1 was a 63-year-old woman referred for a second opinion with a history of nonspecific visual disturbance and a diagnosis of chronic central serous chorioretinopathy complicated by type 1 neovascularization in both eyes. She had received 1 intravitreal injection of aflibercept to the right eye prior to referral. Visual acuities were 20/25 in the right eye and 20/20 in the left eye. Multimodal imaging findings of both eyes are summarized in Figure 1 . Pachychoroid features were seen on indocyanine green angiography and spectral-domain OCT and both fluorescein and indocyanine green angiography were noncontributory for diagnosing neovascularization.
Patient 2 was an 86-year-old white man with pachychoroid neovasculopathy. His right eye had undergone 3 sessions of photodynamic therapy and received 59 anti-VEGF injections; the left eye had received 69 anti-VEGF injections. Visual acuities were 20/70 in the right eye and 20/40 in the left eye and imaging findings are summarized in Figure 2 . Although subfoveal choroidal thickness appeared to be within normal limits, pachyvessels were seen on cross-sectional and en face OCT with thinning and loss of the overlying choriocapillaris and Sattler layers. Extrafoveal choroidal thickness measurements were significantly greater where pachyvessels were most densely located. Dye angiography (not shown) was noncontributory.
Patient 3 was a 70-year-old African-American woman with polypoidal choroidal vasculopathy. Findings are summarized in Figure 3 . She had presented at baseline with acute visual loss and was found to have extensive subretinal and sub-RPE hemorrhage with a recent subfoveal bleed. She had subsequently undergone 3 sessions of photodynamic therapy and received 62 anti-VEGF injections in the right eye. By year 6, corrected visual acuity of the study eye was 20/70, improving to 20/30 with pinhole, and remained stable thereafter. At that time spectral-domain OCT showed a shallow irregular PED. At the last follow-up in year 8, clinical and OCT findings were unchanged and OCT angiography through the PED showed subfoveal type 1 neovascularization with mature feeder vessels extending from the fovea toward the peripapillary site of the polypoidal lesions.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
