Optical Coherence Tomography Angiography to Assess Pigment Epithelial Detachment
Retinal pigment epithelial detachment (PED) is an anatomical separation of the retinal pigment epithelium (RPE) from the underlying Bruch’s membrane.1 PEDs are most commonly associated with age-related macular degeneration (AMD), but are also seen in other retinal disorders, such as central serous chorioretinopathy (CSCR) and neoplastic, inflammatory, or iatrogenic conditions. In AMD, PEDs can be categorized into drusenoid, serous, vascularized, or mixed categories. Vascularized PEDs are associated with type 1 (sub-RPE) choroidal neovascularization (CNV) and typically portend a higher risk for vision loss than other types of PED.
Traditionally, fluorescein angiography (FA) has been the gold standard for diagnosing and assessing PEDs. However, FA is invasive, time consuming, provides only a two-dimensional (2D) image, and has a minimal but significant risk profile that includes nausea, allergy, and, rarely, anaphylaxis. Additionally, FA images can be obscured by dye leakage. These limitations reduce the utility of FA in specific patient populations and represent obstacles to the use of FA as a primary screening instrument.
Optical coherence tomography angiography (OCTA) is a noninvasive, depth-resolved, non-dye-based, and rapid technique for visualizing retinal and choroidal vasculature, as well as various planes of retinal tissues. OCTA is capable of providing en-face, detailed, three-dimensional (3D) images of PEDs and CNV. Recent studies have demonstrated the effectiveness of OCTA in identifying and evaluating CNV even in eyes in which CNV was not visualized by other imaging modalities.2,3 The purpose of the study reported in this article is to describe the spectrum of retinal findings associated with the various types of PEDs through the use of OCTA without the need for invasive dyes.
Images were obtained using the AngioVue OCTA system operating at 70,000 A-scans per second, using a light source centered on 840 nm and a bandwidth of 50 nm. OCTA volumes contained 304 x 304 A-scans with two consecutive B-scans that were captured at each fixed position before proceeding to the next sampling location. Split-spectrum amplitude-decorrelation angiography (SSADA) was used to extract the OCTA information.2 Each OCTA volume was acquired over 3 seconds, and two orthogonal OCTA volumes were acquired to perform motion correction to minimize the motion artifacts arising from microsaccades and fixation changes.
The angiography information is displayed as the average of the decorrelation values when viewed perpendicularly through the thickness being evaluated. The modifications in reflectivity are directly related to blood flow. The horizontal and vertical scans were combined with a motion correction technology (MCT) algorithm that compensates for the motion of the patient’s eyes to create a 3D volume assessment of the retinal vascularization. Qualitative analysis and comparisons of the entire imaging data set were conducted in 3 to 5 minutes. OCTA angiograms were correlated with concurrently obtained high-resolution spectral domain optical coherence tomography (SD-OCT) B-scans. This allowed for visualization of both retinal structure and blood flow, which was useful for identifying subretinal fluid and other anatomic nonvascular features relevant in the diagnosis and assessment of PED.
The OCTA software was used to delineate a region of interest with an inner border at the level of the outer aspect of the inner nuclear layer (seen as a green line on the corresponding SD-OCT B-scan) and an outer border at the level of Bruch’s membrane (observed as a red line on the corresponding SD-OCT B-scan). Additionally, OCTA images were segmented manually into four layers: superficial and deep plexi (to show retinal vasculature), outer retina (to identify CNV), and choriocapillaris. Specific desired locations of retinal segmentation were identified by meticulously scrolling through OCTA images and their corresponding OCT B-scans. The outer border of each segment was then individually adjusted to align with Bruch’s membrane.
PEDs were visualized on OCTA in 44 eyes of 44 patients. Patient age ranged from 45 to 70 years, with a mean age of 68.7 ± 11.5 years. All patients were Caucasian and 24 (54.5%) were women. Of the 44 eyes, four had drusenoid PED (9%), 10 had serous PED (22.7%), 28 had vascularized PED (vPED) (63.6%), and two had mixed PED (4.6%). No eye had more than one type of PED. In the 10 eyes with serous PEDs and four eyes with drusenoid PEDs, OCTA discerned PED without neovascularization. In all 28 eyes with vPEDs and in both eyes with mixed PED, OCTA imaging identified CNV. With the SD-OCT B-scan, we utilized the built-in caliper to manually measure the length and thickness of vPEDs from the inner border of the RPE (identified as a hyperreflective line) to the inner surface of the retina. Mean vPED length was 2,469 μm ± 1,182 μm and mean thickness was 203 μm ± 69 μm. We classified the 28 vPEDs into two distinct subtypes according to features identified on OCT: flat irregular vPEDs (thickness of 0 μm to 160 μm) and dome-shaped vPEDs (thickness of 160 μm to 450 μm).
This study using OCTA to assess specific vascular and nonvascular features of PED in patients with AMD and other retinal pathology demonstrates that OCTA imaging is capable of differentiating between nonvascular and vascularized PEDs, and that OCTA can also assess and measure CNV in vPED. Similar to prior studies, we found that OCTA is capable of visualizing PED via semiautomated segmentation of the outer retina and subretinal or sub-RPE space. Additionally, our findings support prior reports in showing that OCTA imaging, along with concurrently obtained OCT B-scans, is useful in identifying subretinal fluid and other anatomic nonvascular features relevant in the assessment of PED.
Using SD-OCT, we classified vPEDs into two subtypes according to features identified on imaging: flat irregular vPED and dome-shaped vPED. After assessing OCTA images and corresponding SD-OCT B-scans of vPED, we observed that CNV structures in flat irregular vPEDs appeared monolayered on OCTA, whereas CNV in dome-shaped vPEDs appeared multilayered on OCTA. To analyze dome-shaped vPED using OCTA, multiscan analysis was necessary to highlight the entire structure of the neovascular network.
A novel method of accurate OCTA imaging segmentation was utilized in this study. This process entailed meticulously scrolling through OCTA images and their corresponding OCT B-scans to pinpoint specific desired locations of retinal segmentation, thus allowing us to determine vasculature depth. This process, although time consuming, was helpful in differentiating CNV from choroidal vessels, especially when determining the boundaries of CNV. In a busy clinic setting, however, dynamic adjustment of OCTA boundaries is not a feasible option. Development of an improved OCTA algorithm capable of accurate automatic retinal segmentation would benefit clinicians, allowing for more precise noninvasive visualization of vPED and CNV.
OCTA evaluation of PEDs is not without limitations. As noted above, the process used to assess PEDs in this study is time consuming. In addition, patients must be cooperative and have good fixation ability. We used a Spectralis OCT instead of OCTA to calculate PED thickness and length, because OCTA has relatively low resolution and doesn’t permit the use of a caliper to measure directly on the image.
In conclusion, this is the first study that demonstrates the ability of OCTA to noninvasively analyze and differentiate PED into subtypes, as well as detect CNV and subretinal fluid noninvasively. OCTA identified details of CNV structures that appeared monolayered in flat irregular vPEDs and multilayered in dome-shaped vPEDs. In order to analyze dome-shaped vPEDs, a multiscan analysis to highlight the entire structure of the neovascular network is necessary.
1. Mrejen S, Sarraf D, Mukkamala SK, Freund KB. Multimodal imaging of pigment epithelial detachment: a guide to evaluation. Retina. 2013;33(9):1735-1762.
2. Quaranta-El Maftouhi MQ, El Maftouhi A, Eandi CM. Chronic central serous chorioretinopathy imaged by optical coherence tomographic angiography. Am J Ophthalmol. 2015;160(3):581-587.
3. Jia Y, Bailey ST, Wilson DJ, et al. Quantitative optical coherence tomography angiography of choroidal neovascularization in age-related macular degeneration. Ophthalmology. 2014;121(7):1435-1444.