Pigment Epithelial Detachment
Retinal pigment epithelial detachments (PEDs) are structural splitting within the inner aspect of Bruch’s membrane separating the retinal pigment epithelium (RPE) from the remaining Bruch’s membrane.
Retinal pigment epithelial detachments (PEDs) are characterized by separation between the RPE and the inner most aspect of Bruch's membrane. The space created by this separation is occupied by blood, serous exudate, drusenoid material, fibrovascular tissue or a combination of the above.
Various ocular and systemic diseases can be associated with PEDs, the most prevalent being age-related macular degeneration (both dry and wet forms). PEDs can also be present in several chorioretinal diseases including idiopathic central serous chorioretinopathy (CSC) small multifocal idiopathic PEDs, polypoidal choroidal vasculopathy (PCV), and retinal angiomatous proliferation. Rare cases of Vogt Koyanagi Harada syndrome (VKH)-associated PEDs have been reported.
PEDs have also been associated with certain systemic conditions including renal (tubulo-interstitial nephritis and uveitis syndrome and type II membranoproliferative glomerulonephritis), inflammatory (systemic lupus erythematosis, inflammatory bowel disease, sarcoidosis), infectious (Blastocystis hominis, poststreptococcal syndrome, neurosyphilis), neoplastic (paraproteinemias including cryoglobulinemia, IgA or IgM gammopathies), Waldenström macroglobulinemia, large cell non–Hodgkin lymphoma (ocular–central nervous system form), choroidal nevi, acute myeloid leukemia, and Iatrogenic reasons (Pamidronate, hemodialysis, organ transplantation.)  
The classification of PEDs in AMD can be divided based on their contents. These include drusenoid PED, serous PED, fibrovascular PED, or mixed components. Drusenoid PEDs are seen mostly in non-neovascular or dry AMD. Serous PEDs are typically associated with the neovascular or wet form of AMD, but their natural history is relatively more favorable. Vascularized PEDs associated with Type 1 (sub-RPE) neovascularization and wet AMD, in contrast, have a greater risk of vision loss. Fibrovascular PED is a subset of occult CNV. In eyes with AMD, it is not uncommon to see more than one type of PED.
The development of choroidal neovascularization (CNV) has been associated with long-standing PEDs of all forms. Recognition of its existence is a major concern secondary to its increased risk for severe vision loss.
This article largely focuses on PED in the setting of AMD. Features of PED in idiopathic polypoidal choroidal vasculopathy are described in detail here.
The retinal pigment epithelium consists of a monolayer of cells, attached to its overlying retinal photoreceptors apically, and the inner collagenous layer of Bruch’s membrane on its basal surface. Proper anatomical apposition between the retina, the RPE, and Bruch’s membrane is crucial for proper retinal function, such as retinal metabolism, nutritional support of photoreceptors, phagocytosis of photoreceptor outer segments, and formation of the outer blood-retinal barrier. The inner surface of the RPE cells contain microvilli that interdigitate with the photoreceptor outer segments to stabilize this apposition. Such interdigitations of microvilli do not exist between RPE and Bruch’s, but these layers are possibly thought to remain attached together through a combination of hydrostatic and osmotic forces from the vitreous as well as possible regions of hemi-desmosomes.
A PED occurs when the basal lamina of the RPE cell separates from the inner collagenous Bruch’s membrane. There are several ongoing hypotheses regarding the pathophysiology of PED. Age-related deposition of lipids, such as cholesterol esters, triglycerides, and fatty acids in Bruch's membrane may thicken and change its permeability altering retinochoroidal flow. Fluid may accumulate in the sub- RPE space, unable to pass through Bruch’s membrane, resulting in RPE elevation and a PED. The development of drusen may disrupt fluid outflow through Bruch’s membrane in a similar fashion.
In rarer inflammatory etiologies of PED, choroidal inflammation might increase vascular permeability and breakdown of the outer blood-retinal barrier leading to protein-rich fluid accumulating under the RPE. Ischemic processes may similarly cause endovascular damage of choroidal and retinal vessels and resulting blood-retinal barrier breakdown.
Diagnosing PEDs relies on careful physical exam and further evaluation with various imaging modalities.
PEDs are often asymptomatic, but patients can present with painless blurred vision, distortion, metamorphopsia, or induced hyperopia.
Clinical Exam and Diagnostic Imaging
PEDs appear as single or multiple elevated mounds on fundus examination, with sizes ranging from sub-biomicroscopic to several disc diameters. Pigmentary changes such as clumping or mottling may be present. Orange pigment rings and bands on the dome, extending radially over the PED, may indicate chronicity of disease. PEDs are better appreciated in retroillumination during slit lamp biomicroscopy with 90D or 78D lens. The development of high-resolution spectral domain (SD) OCT has allowed more precise and comprehensive assessment and longitudinal follow up on PEDs. 
Examination: Drusenoid PEDs appear as well-circumscribed yellow or yellow–white elevations of the RPE that are usually found within the macula. They may have scalloped borders and a slightly irregular surface. It is not uncommon to observe a speckled or stellate pattern of brown or gray pigmentation on their surface.
OCT: Drusenoid PEDs usually show a smooth contour of the detached hyperreflective RPE band that may demonstrate an undulating appearance. The material beneath the RPE band typically exhibits a dense homogeneous appearance with moderate or high hyperreflectivity. Drusenoid PEDs are typically not associated with overlying subretinal or intraretinal fluid. (Figure 1A, arrow)
Fundus autofluorescence (FAF): Drusenoid PEDs may exhibit decreased FAF but typically they are isofluorescent or hyperautofluorescent. Drusenoid PEDs often show an evenly distributed, modest increase in the FAF signal surrounded by a well defined, hypoautofluorescent halo delineating the entire border of the lesion.
Fluorescein Angiography (FA): Drusenoid PEDs demonstrate faint hyperfluorescence in the early phase that increases throughout the transit stage of the study without late leakage. This gradual staining without leakage may be sprinkled by hypofluorescent foci corresponding to pigment deposits.
Indocyanine green angiography (ICG): The content of the drusenoid PED will block the fluorescence emitted from the underlying choroidal vasculature and, therefore, the PED will appear as a homogeneous hypofluorescent lesion during the early phase and remain hypofluorescent throughout the transit.
Examination: Serous PED appears as a distinct circular or oval-like detachment of the RPE. Clear or yellowish–orange in color, this dome-shaped elevation of the RPE has a sharply demarcated border. Often PEDs will transilluminate if they are filled predominantly with serous fluid when observed at the slit lamp.
OCT: On OCT, serous PEDs appear as well-demarcated, abrupt elevations of the RPE with a homogenously hyporeflective sub-RPE space. Enhanced depth imaging (EDI) OCT is useful to determine whether serous PED is caused by AMD (normal subfoveal choroidal thickness) or by CSC (increased subfoveal choroidal thickness). (Figure 1B)
FAF: Serous PEDs most often have an even distribution of hyperautofluorescence corresponding to the detachment and are surrounded by a hypoautofluorescent border.
FA: Serous PEDs demonstrate intense early hyperfluorescence and brisk, progressive pooling within the PED in a homogeneous and well-demarcated manner. Late staining of serous PEDs is typical and may make it difficult to differentiate these PEDs from those that are vascularized based on FA alone. Serous PED may obscure the boundaries / extent of an associated CNV. In cases where there is suspicion of associated CNV, ICGA is a useful imaging modality.
ICG: ICGA will typically demonstrate round hypofluorescence with with sharply lineated borders – late stages remain typically hypofluorescent as well.
Examination: Gass reported that a flattened or notched border of the PED is a frequent and important sign of hidden associated CNV. Other biomicroscopic findings suggestive of possible occult CNV association include yellow subretinal and intraretinal exudates that occur typically at the PED margins, subretinal hemorrhages at PED margins, sub-RPE blood which appears darker than subretinal blood with a fluid-level sign, irregular elevation of the PED because of organization in the lesser elevated area, and radial chorioretinal folds surrounding the PED caused by the contraction of Bruch membrane and CNV.
OCT: OCT demonstrates irregular RPE elevation in contrast to the smooth elevation of a serous PED, with heterogenous features often along the back surface of the detached RPE. Scant overlying subretinal fluid may sometimes be seen. (Figure 1C, arrow)
FAF: Fundus autofluorescence imaging of vascularized PEDs has not been evaluated systematically in large series of patients. More work needs to be done to correlate the FAF pattern of PEDs and any associated CNV with findings obtained with FA and OCT.
FA: Stippled or granular hyperfluorescence (not as bright as classic CNV) and usually appears within 1-2 minutes. In later frames, these areas can intensify and demonstrate persistent staining or leaking. Other features include tiny focal hyperfluorescent pinpoint spots in mid- to late-phase frames that do not correspond to drusen or foci of depigmentation. An additional fluorescein angiographic pattern of vascularized PEDs is a serous PED with a notch (e.g., kidney bean–shaped PED) or hot spot that may be referred to as a vascularized serous PED.
ICG: Leakage of ICG in the late phases or focal “hot spots” may be indicative of an underlying associated CNV. A focal bright area of well-defined hyperfluorescence less than 1 disk diameter in size referred to as a hot spot or focal CNV. Plaque CNV on the other hand, occupies a large portion of the PED.
Depending on the etiology of the PED, different treatment modalities have been explored to prevent vision loss.
Currently no treatment for serous PED is proven effective, nor are recommendations for treatment guidelines established. Several strategies, however, have been used to treat vascularized PEDs, including laser photocoagulation, photodynamic therapy (PDT), intravitreal steroids and anti-VEGF therapy. The results from the VIP trial indicated that PDT could significantly reduce the risk of moderate and severe vision loss among patients with subfoveal occult CNV. Another treatment modality, described recently by Costa et al as a pilot trial, is photothrombosis at the neovascular ingrowth site using ICG visualization followed by laser application to feeder vessels. Occlusion of the feeder vessel with cessation of leakage, restoration of macular architecture and visual improvement were induced in two patients with CNV associated with PEDs. 
The natural history of PEDs vary by type, but any PED may be complicated by sight-threatening RPE tears. The most important FA feature identified which predicted RPE tears was the uneven filling of the PED, with a hypofluorescent central area that remained dark until the late angiographic frames as well as early hyperfluorescence at the borders of the PED that grew progressively, and sometimes demonstrated a notched edge.  The RPE tear rate in eyes with PED in natural history studies has been noted to be between 10% and 12%, but this rate seems to be accelerated after anti-VEGF therapy (up to 17%). Other complications include the development of CNV (occult or classic), and geographic atrophy.
Drusenoid PEDs carry the best visual prognosis. Additionally, the location of the PED is important in determining prognosis. Patients with extrafoveal PEDs tend to preserve good visual acuity, whereas patients with subfoveal PEDs can have worse visual outcomes. The course of PEDs also varies by disease etiology. Mudvari et al demonstrated with a mean follow-up of 49 months that 65% of PEDs in CSC completely resolved and the other 35% PEDs remained persistent. Retinal pigment epithelium atrophy was evident in 86% of patients over the area of the resolved PED. Others have reported similar findings and noted that patients with a persistent PED had poorer visual outcomes. 
Some patients with fibrovascular PED can appear relatively asymptomatic and may never experience vision loss despite continued growth of the neovascular lesion. Larger vascularized or hemorrhagic PEDs are typically associated with significant vision loss.
- American Academy of Ophthalmology. Pigment epithelial detachment. https://www.aao.org/image/pigment-epithelial-detachment-2 Accessed July 05, 2019.
- Khochtali S, Ksiaa I, Megzari K, Khairallah M. Retinal Pigment Epithelium Detachment in Acute Vogt-Koyanagi-Harada Disease: An Unusual Finding at Presentation. Ocul Immunol Inflamm. 2019;27(4):591–594.
- Mrejen, S. (2013). Multimodal imaging of pigment epithelial detachment: a guide to evaluation. Retina, 33(9), 1735-1762.
- Wolfensberger, Thomas J., and Adnan Tufail. "Systemic disorders associated with detachment of the neurosensory retina and retinal pigment epithelium." Current opinion in ophthalmology 11.6 (2000): 455-461.
- Zayit-Soudry, S., Moroz, I., & Loewenstein, A. (2007). Retinal pigment epithelial detachment. Survey of ophthalmology, 52(3), 227-243
- Green K, Paterson CA, Cheeks L, et al: Ocular blood flow and vascular permeability in endotoxin-induced inflamma- tion. Ophthalmic Res 22:287--94, 1990
- Starita C, Hussain AA, Patmore A, et al: Localization of the site of major resistance to fluid transport in Bruch’s membrane. Invest Ophthalmol Vis Sci 38:762--7, 1997
- Yannuzzi LA: Retinal pigment epithelial detachment, in Yannuzzi LA (ed): Laser Photocoagulation of the Macula. Philadelphia, Lippincott, 1989, pp 49--63
- Lee SY, Stetson PF, Ruiz-Garcia H, Heussen FM, Sadda SR. Automated characterization of pigment epithelial detachment by optical coherence tomography. Invest Ophthalmol Vis Sci. 2012 Jan 20;53(1):164-70.
- Arnold, J. J., Quaranta, M., Soubrane, G., Sarks, S. H., & Coscas, G. (1997). Indocyanine green angiography of drusen. American journal of ophthalmology, 124(3), 344-356.
- Imamura, Y., Fujiwara, T., Margolis, R. O. N., & Spaide, R. F. (2009). Enhanced depth imaging optical coherence tomography of the choroid in central serous chorioretinopathy. Retina, 29(10), 1469-1473.
- Gass, J. D. (1984). Serous retinal pigment epithelial detachment with a notch. A sign of occult choroidal neovascularization. Retina (Philadelphia, Pa.), 4(4), 205-220.
- Arnold, J., Barbezetto, I., Birngruber, R., Bressler, N. M., Bressler, S. B., Donati, G., ... & Kaiser, P. K. (2001). Verteporfin therapy of subfoveal choroidal neovascularization in age-related macular degeneration: two-year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization-verteporfin in photodynamic therapy report 2. American journal of ophthalmology, 131(5), 541-560.
- Costa, R. A., Rocha, K. M., Calucci, D., Cardillo, J. A., & Farah, M. E. (2003). Neovascular ingrowth site photothrombosis in choroidal neovascularization associated with retinal pigment epithelial detachment. Graefe's archive for clinical and experimental ophthalmology, 241(3), 245-250.
- Coscas, G., Koenig, F., & Soubrane, G. (1990). The pretear characteristics of pigment epithelial detachments: a study of 40 eyes. Archives of ophthalmology, 108(12), 1687-1693.
- Mudvari, S. S., Goff, M. J., Fu, A. D., McDONALD, H. R., Johnson, R. N., Ai, E., & Jumper, J. M. (2007). The natural history of pigment epithelial detachment associated with central serous chorioretinopathy. Retina, 27(9), 1168-1173.
- Loo, R. H., Scott, I. U., FLYNN Jr, H. W., Gass, J. D. M., Murray, T. G., Lewis, M. L., ... & Smiddy, W. E. (2002). Factors associated with reduced visual acuity during long-term follow-up of patients with idiopathic central serous chorioretinopathy. Retina, 22(1), 19-24.
- Blinder, K. J., Bradley, S., Bressler, N. M., Bressler, S. B., Donati, G., Hao, Y., ... & Pournaras, C. (2003). Effect of lesion size, visual acuity, and lesion composition on visual acuity change with and without verteporfin therapy for choroidal neovascularization secondary to age-related macular degeneration: TAP and VIP report No. 1. American journal of ophthalmology, 136(3), 407-418.