This study shows features in the development of Type 3 neovascularization that expands the underlying concepts of disease pathophysiology. Edema and hemorrhage were found to occur before neovascularization. In eyes with neovascularization, there were widespread edema and hemorrhage, much greater in extent than the area of neovascularization, and were not necessarily contiguous with the new vessels. With evolution of neovascularization, there appeared to be rapid vascular remodeling of vessels associated with the lesion, including vessels supporting the lesion involving the superficial plexus. Treatment brought a rapid reduction in the amount of edema and hemorrhage and associated retinal vascular changes. Although these seem to be simple findings, they differ from what has been previously reported and they have significant implications in our understanding of Type 3 disease. To fully understand these implications, it is important to appreciate historical developments in the perception of the disease, particularly the contradictory nature of what has been reported over the past two and a half decades.
In 1992, Harnett et al3 reported unusual retinal pigment epithelial detachments that were associated with retinal vessels that dove down into the deep retina and formed an angiomatous lesion. In 1996, Harnett et al4 further defined what they called “deep retinal vascular anomalous complexes” as having at least one retinal arteriole dipping to a deep vascular complex and at least one draining retinal venule with an intervening angiomatous lesion. The angiographic examples showed the vascular connections to arise from the inner retina. The authors described additional characteristics, including blot and nerve fiber layer intraretinal hemorrhages, retinal telangiectasis, and microaneurysms. They proposed that outer retinal “hypoxia, angiogenic factors such as vascular endothelial growth factor released by Müller and other cells may encourage retinal vessels to proliferate in a directional manner toward the outer retina.”
In 2000, Lafaut et al10 reported the histopathologic findings of six specimens from eyes with deep retinal anomalous complexes that were obtained at the time of retinal translocation surgery. Four of the cases had angiography and showed retinal vascular vessels from the inner retina descending into the retina to the deeper angiomatous lesion. None of the cases showed vascular connections to the choroid by histology and two specimens had retained large retinal vessels attached to the excised lesion.10
In 2001, Yannuzzi et al5 published a paper covering the clinical aspects of 143 eyes of 108 patients with an entity that they called “retinal angiomatous proliferation” (RAP). The neovascularization appeared to be identical to that reported earlier by Hartnett et al, but Yannuzzi et al fleshed out what they considered to be the pathophysiologic development of neovascularization. They proposed that neovascularization originated from the deep vascular plexus and grew toward the superficial retina and deep retina simultaneously. This was Stage I RAP (see Figure 1 of the article by Yannuzzi et al). In Stage II RAP, the newly proliferating vessels extended out of the retina into the vitreous and were associated with preretinal hemorrhages. There was also a large vascular structure in the inner retina in the schematic drawings, but this structure was not defined. In addition to anterior retinal neovascularization, vessels proliferated posteriorly toward the RPE. The vessels originating from the neovascularization were capillaries, but they did not change in caliper after the development of the new vessels. The new vessels appeared to be smaller in diameter than the capillaries. In both the inner and outer neovascularization, there were dozens of new retinal vessels drawn in the schematics. Curiously, the accompanying angiographic images showed a limited number of vessels, sometimes as few as two, descending into the retina. No angiographic example of preretinal neovascularization was shown. The schematics showed cystoid collections of fluid in the retina were contiguous with individual newly growing vessels.
In 2003, Spaide11 reported a significant association of focal hyperpigmentation with RAP in the fellow eye; both of these conditions showed bilateral symmetry. Spaide et al12 proposed the focal hyperpigmentation was caused by detached retinal pigment epithelial cells that migrated into the retina but retained their ability to secrete VEGF. This caused a VEGF gradient that helped induce the posterior growth of vessels in the disease. In 2003, Gass et al6 proposed a different explanation than what was proposed by Hartnett et al4 and Yannuzzi et al.5 Gass et al believed there was preexisting occult neovascularization with coexistent atrophy of the outer retina. With subsidence of the retina, the retinal blood vessels came near the occult neovascularization, which eventually resulted in chorioretinal anastomoses and associated intraretinal hemorrhage. The superficial hemorrhage was a sign of occult chorioretinal anastomosis in occult choroidal neovascularization, thus two uses of the word occult in a one-phrase disease description.
Shimada et al13 reported the histopathologic findings of specimens obtained by surgical excision of RAP in nine eyes of eight patients in 2007. The neovascular masses expressed VEGF and included macrophages and RPE cells. Hypoxia-inducible factors (HIF-1 alpha and HIF-2 alpha) were also expressed.13 Histologic examination by Monson et al, and Klein and Wilson showed prominent intraretinal hemorrhages could be found in the superficial retina, even the nerve fiber layer, well away from the neovascularization. Intraretinal RPE cells were observed to be contiguous with the neovascular lesion.14,15 Like the Lafaut et al's paper, none of these publications showed any vascular connection to the choroid.
In 2008, Krebs et al16 had a schematic progression of RAP (shown in Figure 1 of the article), and like Yannuzzi et al, they showed the vessels started in the deep plexus and proliferated anteriorly and posteriorly in the retina. The newly growing vessels were proposed to penetrate the internal limiting membrane to enter the vitreous cavity anteriorly, whereas the posteriorly proliferating vessels reached and then penetrated the RPE layer. There were hundreds of proliferating vessels shown. The two supporting angiographic cases shown in Figures 2 and 3 of the article, nevertheless, showed two vessels connecting the superficial retinal plexus to an angiomatous lesion in one figure and no visible connection from any layer vessels in the other, respectively.
In 2008, Freund et al reported five eyes of four patients with early Stage 1 RAP lesions. They stated the analysis of selected cases suggested a choroidal origin of the neovascular complex with early retinal choroidal anastomosis and evidence of early Type 1 neovascularization.7 They did not state how this would be differentiated from a mixed Type and Type 2 lesion. They specifically stated that because RAP may originate not only from deep retinal capillaries but also from the choroid, the term “Type 3 neovascularization” should be used instead of RAP. They updated the drawings from Yannuzzi et al's three-stage classification to include the theory of choroidal origin while retaining the preretinal neovascularization and large superficial intraretinal vessels. They showed the cystoid edema and individual hemorrhages being closely associated individual vascular segments in the neovascular growth. The new vessels were shown to arise from the capillaries in the retina. With the advent of neovascularization, the originating vessels did not change in diameter and the new vessels were smaller in diameter than the capillaries.
Querques et al17 reported in 2013 that the precursors to Type 3 neovascularization were baseline hyperautofluorescence that turned to focal hypoautofluorescence over time. They also reported that there was localized RPE elevation with focal disruption of the RPE and photoreceptors and “the overlying outer plexiform layer that progressively took contact with the RPE.” It is not clear if the contact was with the RPE or the RPE defect.17 In the same year, Querques et al18 reported 19 eyes of 19 patients with Type 3 neovascularization with simultaneous OCT and angiography. They stated all eyes had focal discontinuity of the RPE monolayer, there was a single retinal arteriole feeding into the lesion, and choroidal contributions to the neovascular process could not be visualized.
In 2015, Kuehlewein et al19 reported the OCT angiographic findings of Type 3 neovascularization in 29 eyes of 24 patients. They believed the neovascular complex appeared as a “small tuft of bright, high-flow tiny vessels with curvilinear morphology located in the outer retinal layers with a feeder vessel communicating with the inner retinal circulation (i.e., deep retinal capillary plexus).” This is difficult to understand because the deep capillary plexus is separate from the inner, or superficial, vascular circulation. Of the 29 eyes, 16 were active and in 10 of these vessels could be visualized, whereas in the 13 inactive eyes no vessels could be visualized. In 2015, Miere et al20 reported 18 treatment-naive eyes of 18 consecutive patients; OCT angiography reportedly showed lesions characterized by retinal–retinal anastomosis that emerged from the deep capillary plexus and descended down to abut the RPE.
In 2015, Querques et al21 (the same authors as Ref. 18) published an editorial, which also contained supporting clinical imaging information. They stated that “in many cases, the early appearance of Type 3 neovascularization is characterized by an intraretinal vascular complex emanating from the deep capillary plexus often associated with adjoining telangiectatic vessels.” The lesions were said to be retinal–retinal anastomoses. In their schematics, they showed Type 3 neovascularization arising as a multitude of vessels (unlike their 2013 paper)18 from the deep capillary plexus. The drawings did not show extension of the neovascularization into the inner retina until late in the disease, but when it happened, the extension was by numerous vessels. This suggests there must be a VEGF gradient to the inner retina. Cystoid spaces were shown to be directly contiguous with the new vessel elements. Unlike the 2013 paper, many cases shown did not have subsidence of the inner and middle retina, and high-resolution OCT scanning did not show any apparent defect in the RPE. The authors retained the idea that Type 3 neovascularization could arise from the choroid, in a process where many vessels broke through the RPE and connected to the deep plexus with no change in the diameter of the deep plexus capillaries.21
Su et al22 reported 34 eyes with new-onset Type 3 neovascularization in 2016. Some eyes had what the authors considered to be a preonset OCT; the majority of these, 77.8%, had hyperreflective foci. With initiation of neovascularization, which grew toward the deep retina, there was also neighboring cystoid edema in the retina. As the neovascularization descended, the associated cystoid spaces became larger. The neovascularization originated from the deep vascular plexus and did not involve any more superficial layers with maturation of the lesion, which contained what appeared to be dozens of small vessels (see Figure 1 in Su et al22). They showed one angiographic image, which demonstrated two vessels from the superficial vascular plexus descending to a deeper lesion and no apparent deeper vascular contributions. In 2018, Sacconi et al23 reported a condition they termed “nascent Type 3 neovascularization.” Hyperreflective foci were found to have flow signal, suggesting to the authors these foci, as opposed to previous proposals, may represent new vessels. The authors did not use projection artifact removal in their imaging. They stated the neovascularization did not show leakage until the vessels progressed into the RPE and sub-RPE space.23
There are many common elements to previously reported schema for the development of Type 3 neovascularization. The first steps involve vessels originating from the deep plexus. The main vector for growth of these vessels appeared to be posteriorly, in which a diffusely distributed multitude of vessels, dozens in number, descended toward the RPE. The deep vascular plexus is composed of capillary-sized vessels,24 but each of these capillaries was shown to sprout many smaller-sized vessels and so were vessels smaller than capillaries. In none of the published models do the vessels in the deep plexus enlarge after this occurs, so it is difficult to know how the increased blood flow of the lesion was delivered. A choroidal connection or origin is shown in these schemas, and the many capillaries proposed to be involved sprout from a few capillaries in the choriocapillaris. The sprouting vessels were a fraction of the diameter of the choroidal capillaries, and the choriocapillaries do not change in diameter once this occurs (note that vessels have a lower limit in terms of size, given the dimensions of the cellular elements in blood). Either initially or in some models later, a myriad of small vessels grew into and connected with the superficial vascular plexus, but the inner vessels did not change in diameter. In addition, the direction of growth appears to be directed anteriorly, which is running counter to the vector of the previously growing vessels. In several papers, vessels are shown to enter the vitreous cavity, but corresponding clinical evidence for this has not been published. Any associated hemorrhage or cystoid spaces were shown to be contiguous with specific vessels in the neovascular complex. These models did not agree with the observed vascular growth contemporaneously imaged with fluorescein and indocyanine green angiography, which generally showed only one or a few feeding and draining vessels that connected directly to the superficial vascular plexus. These models did not agree with contemporaneous histopathologic studies.
Tolentino et al injected VEGF into the vitreous cavity of monkeys.25–27 Elevated levels of VEGF were found to cause retinal vascular leakage, hemorrhage, and telangiectasis. Curiously, most eyes did not develop retinal neovascularization, but a few did produce minimal peripheral neovascularization.25,26 Vascular endothelial growth factor helps initiate the early stages of angiogenesis in which there is vascular leakage and proliferation of capillary tubes.28 From animal models of ischemia and data obtained from treating humans, we know VEGF blockade is effective in causing neovascularization to regress.29–31 This implies VEGF is necessary, but may not be sufficient in all disease states, to cause pathologic retinal neovascularization, although it is sufficient to cause iris neovascularization.27 We see VEGF-driven leakage and edema in diabetic retinopathy and veno-occlusive disease, but these patients do not develop intraretinal neovascularization in the macula. This suggests that in diseases where vessels do grow, such as Type 3 neovascularization, there must be a permissive local environment. If there is enough VEGF in the retina to elicit neovascularization, there is reason to expect vascular leakage, telangiectasis and hemorrhage would also occur independent of the new vessel growth. The new vessels also have the potential to contribute to the observed edema and hemorrhage.
Vascular remodeling is expected to occur with demands to deliver blood flow to the newly growing complex, with enlargement of a limited number of retinal vessels feeding and draining the lesion. Vascular enlargement in vessels is dependent on shear stress on the vascular endothelial cells and is not dependent on VEGF.32 Vascular enlargement can occur in newly proliferated or in established vessels. Ultimately the flow in the lesion is derived from the superficial vascular plexus. Efficient remodeling would create feeder and draining vessels connecting to the superficial plexus instead of dozens of small vessels connecting to the deep plexus, which in turn would require some method of conducting increased blood flow to and from the superficial plexus. Treatment with anti-VEGF agents causes a regression of the newly growing vascular sprouts. It is possible that older, arteriolized components of the neovascularization may not regress with VEGF withdrawal. Even so, if the lesion is sufficiently suppressed, there could be vascular remodeling leading to vascular regression. Histopathologic evaluation has shown there is a vascular network of small capillary-sized vessels in the angiomatous proliferation between the larger feeding and draining vessels, as would be expected in an acquired angiomatous growth of vessels proliferating because of a stimulus. Therefore, the term “retinal–retinal anastomosis” is not accurate as a primary descriptor for Type 3 disease.
The source of the VEGF is likely to be multifactorial. The choriocapillaris is attenuated under areas of drusen and subretinal drusenoid deposits.33 Optical coherence tomography angiography shows signal voids in eyes with pseudodrusen and drusen.34,35 This loss could lead to outer retinal ischemia. Estimates of the oxygen diffusion over drusen are limited by the lack of knowledge of the diffusion coefficient for drusen material. Assuming a solitary drusen 200 µm in diameter, 70 µm in height, and a diffusion coefficient the same as the retina, Linsenmeier and Zhang36 calculated there would be a greater than 40% drop in photoreceptor pO2. In age-related macular degeneration eyes, the choriocapillaris is compromised, there are often confluent drusen and basal laminar deposit, so the actual pO2 reduction is likely to be much worse. Hypoxia may be a factor in causing the RPE cells to detach from the RPE monolayer to migrate up to the more vascularized retina. When the RPE cells depart, one would expect a defect to occur in the monolayer. To what extent the defect persists, how much is filled with Müller cell processes, or is covered by cellular extensions of neighboring RPE cells is not known at present. Inherent in these possibilities is the likelihood that altered diffusion of material, and potentially cytokines, could occur from the choroid. Increased levels of VEGF can contribute to vessel growth, but for it to do so there has to be concomitant changes in cytokine levels produced by other cells, such as Müller cells, which can produce angiogenic and antiangiogenic37–39 cytokines. The total balance of forces exerted by proangiogenic and antiangiogenic forces would ultimately control the proliferation of vessels, but the presence of VEGF, by itself, could cause retinal vascular abnormalities (Figure 11). A gathering storm leads to increased risk of lightning strikes, which may occur in a stochastic fashion. The risk factors associated with the advent of Type 3 neovascularization, such as drusen and focal hyperpigmentation, could lead to a VEGF storm in the retina, potentially leading to increased risk of proliferation and descent of new vessels somewhere in the region involved.
A proposed series of pathophysiologic events is schematically shown in Figure 12. Consistent with previous reports, there are drusen with overlying detachment and migration of RPE cells into the retina. At present, there are not enough data to model how subretinal drusenoid deposits, appearing as pseudodrusen, affect development of disease. Initial stages include release of VEGF leading to edema and small hemorrhages in the retina. Given the lack of visualization of vessels in early cases, it is likely that there is a preproliferative stage in which VEGF-mediated changes occur in the retina before the advent of neovascularization. Therefore, a suggested naming scheme is that the process is Type 3 disease (or alternatively Type 3 vasculopathy) and this stage is called preproliferative Type 3 disease. Over time, in a permissive environment, there is the development of new vessels that grow along a VEGF gradient that increases posteriorly. This phase is called Type 3 neovascularization. Leakage and hemorrhage present could be related to the continued levels of VEGF and from the neovascularization. Vascular remodeling would lead to the evolution of larger feeding and draining vessels, ultimately connecting to the superficial plexus. It is not necessary to posit blood vessel growing into the inner retina, as this would require another VEGF gradient. Inherent in the definition is that Type 3 vessels originate from the retina and not from underlying Type 1 neovascularization or the choroid as was proposed in Refs. 6, 7, and 21.
This study found that areas of hemorrhage and edema were far larger, and not necessarily contiguous, with hyperfluorescence seen during fluorescein angiography in eyes with Type 3 neovascularization. These areas could appear before the advent of actual neovascularization. These findings, in part, may be attributed to increased cytokines, such as VEGF, in the retina because the known physiologic effects of VEGF on the retinal vasculature includes leakage, hemorrhage, and telangiectasis even without the production of neovascularization. These findings mark the preproliferative stage of Type 3 disease. Proliferation and remodeling of associated retinal vasculature is proposed to lead to feeding and draining vessels with an intervening mass of new capillaries. Although the initiation of the neovascularization could be from the deep capillary plexus, after remodeling the connections would likely appear to be from the superficial vascular plexus, as has been shown in many past reports. Lack of prominent fluorescein pooling in the cystoid spaces in the present series is intriguing. In a similar manner, poorly evident pooling of dye in cystoid spaces occurs in retinovascular cystoid macular edema secondary to diabetes or veno-occlusive disease, other forms of macular neovascularization. The various possibilities for this observation could include intraretinal diffusion of fluorescein in the retina occurring at a much slower rate than the aqueous component in edema, the susceptibility to light scattering inherent in fundus camera angiography, and not enough contrast between the amount of dye in the cystoid spaces compared with the surrounding tissue at the time of image acquisition in the angiographic sequence. This study has salient weaknesses, the largest being the limited number of cases evaluated. This has been a weakness of most studies published to date and may be an explanation for the large variability in the findings attributed to Type 3 disease.
1. Gass JD. Pathogenesis of disciform detachment of the neuroepithelium. Am J Ophthalmol 1967;63(Suppl):1–139.
2. Gass JD. Biomicroscopic and histopathologic considerations regarding the feasibility of surgical excision of subfoveal neovascular membranes. Am J Ophthalmol 1994;118:285–298.
3. Hartnett ME, Weiter JJ, Garsd A, Jalkh AE. Classification of retinal pigment epithelial detachments associated with drusen. Graefes Arch Clin Exp Ophthalmol 1992;230:11–19.
4. Hartnett ME, Weiter JJ, Staurenghi G, Elsner AE. Deep retinal vascular anomalous complexes in advanced age-related macular degeneration
. Ophthalmology 1996;103:2042–2053.
5. Yannuzzi LA, Negrão S, Iida T, et al. Retinal angiomatous proliferation in age-related macular degeneration
. Retina 2001;21:416–434.
6. Gass JD, Agarwal A, Lavina AM, Tawansy KA. Focal inner retinal hemorrhages in patients with drusen: an early sign of occult choroidal neovascularization and chorioretinal anastomosis. Retina 2003;23:741–751.
7. Freund KB, Ho IV, Barbazetto IA, et al. Type 3 neovascularization
: the expanded spectrum of retinal angiomatous proliferation. Retina 2008;28:201–211.
8. Spaide RF. Volume-rendered angiographic and structural optical coherence tomography
. Retina 2015;35:2181–2187.
9. Spaide RF, Suzuki M, Yannuzzi LA, et al. Volume-rendered angiographic and structural optical coherence tomography
angiography of macular telangiectasia Type 2. Retina 2017;37:424–435.
10. Lafaut BA, Aisenbrey S, Vanden Broecke C, Bartz-Schmidt KU. Clinicopathological correlation of deep retinal vascular anomalous complex in age related macular degeneration. Br J Ophthalmol 2000;84:1269–1274.
11. Spaide RF. Fundus autofluorescence and age-related macular degeneration
. Ophthalmology 2003;110:392–399.
12. Spaide RF, Armstrong D, Browne R. Continuing medical education review: choroidal neovascularization in age-related macular degeneration
—what is the cause? Retina 2003;23:595–614.
13. Shimada H, Kawamura A, Mori R, Yuzawa M. Clinicopathological findings of retinal angiomatous proliferation. Graefes Arch Clin Exp Ophthalmol 2007;245:295–300.
14. Monson DM, Smith JR, Klein ML, Wilson DJ. Clinicopathologic correlation of retinal angiomatous proliferation. Arch Ophthalmol 2008;126:1664–1668.
15. Klein ML, Wilson DJ. Clinicopathologic correlation of choroidal and retinal neovascular lesions in age-related macular degeneration
. Am J Ophthalmol 2011;151:161–169.
16. Krebs I, Binder S, Stolba U, et al. Subretinal surgery and transplantation of autologous pigment epithelial cells in retinal angiomatous proliferation. Acta Ophthalmol 2008;86:504–509.
17. Querques G, Querques L, Forte R, et al. Precursors of type 3 neovascularization
: a multimodal imaging analysis. Retina 2013;33:1241–1248.
18. Querques G, Souied EH, Freund KB. Multimodal imaging of early stage 1 type 3 neovascularization
with simultaneous eye-tracked spectral-domain optical coherence tomography
and high-speed real-time angiography. Retina 2013;33:1881–1887.
19. Kuehlewein L, Dansingani KK, de Carlo TE, et al. Optical coherence tomography
angiography of Type 3 neovascularization
secondary to age-related macular degeneration
. Retina 2015;35:2229–2235.
20. Miere A, Querques G, Semoun O, et al. Optical coherence tomography
angiography in early Type 3 neovascularization
. Retina 2015;35:2236–2241.
21. Querques G, Souied EH, Freund KB. How has high-resolution multimodal imaging refined our understanding of the vasogenic process in type 3 neovascularization
? Retina 2015;35:603–613.
22. Su D, Lin S, Phasukkijwatana N, et al. An updated staging system of type 3 neovascularization
using spectral domain optical coherence tomography
. Retina 2016;36(Suppl 1):S40–S49.
23. Sacconi R, Sarraf D, Garrity S, et al. Nascent Type 3 neovascularization
in age-related macular degeneration
. Ophthalmol Retina 2018;2:1097–1106.
24. Snodderly DM, Weinhaus RS, Choi JC. Neural-vascular relationships in central retina of macaque monkeys (Macaca fascicularis). J Neurosci 1992;12:1169–1193.
25. Tolentino MJ, Miller JW, Gragoudas ES, et al. Intravitreous injections of vascular endothelial growth factor produce retinal ischemia and microangiopathy in an adult primate. Ophthalmology 1996;103:1820–1828.
26. Tolentino MJ, McLeod DS, Taomoto M, et al. Pathologic features of vascular endothelial growth factor-induced retinopathy in the nonhuman primate. Am J Ophthalmol 2002;133:373–385.
27. Tolentino MJ, Miller JW, Gragoudas ES, et al. Vascular endothelial growth factor is sufficient to produce iris neovascularization and neovascular glaucoma in a nonhuman primate. Arch Ophthalmol 1996;114:964–970.
28. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med 2003;9:669–676.
29. Aiello LP, Pierce EA, Foley ED, et al. Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins. Proc Natl Acad Sci U S A 1995;92:10457–10461.
30. Miller JW, Adamis AP, Aiello LP. Vascular endothelial growth factor in ocular neovascularization and proliferative diabetic retinopathy. Diabetes Metab Rev 1997;13:37–50.
31. Adamis AP, Shima DT, Tolentino MJ, et al. Inhibition of vascular endothelial growth factor prevents retinal ischemia-associated iris neovascularization in a nonhuman primate. Arch Ophthalmol 1996;114:66–71.
32. Spaide RF. Optical coherence tomography
angiography signs of vascular abnormalization with antiangiogenic therapy for choroidal neovascularization. Am J Ophthalmol 2015;160:6–16.
33. Curcio CA, Messinger JD, Sloan KR, et al. Subretinal drusenoid deposits in non-neovascular age-related macular degeneration
: morphology, prevalence, topography, and biogenesis model. Retina 2013;33:265–276.
34. Lane M, Moult EM, Novais EA, et al. Visualizing the choriocapillaris under drusen: comparing 1050-nm swept-source versus 840-nm spectral-domain optical coherence tomography
angiography. Invest Ophthalmol Vis Sci 2016;57:OCT585–OCT590.
35. Spaide RF. Choriocapillaris flow features follow a power law distribution: implications for characterization and mechanisms of disease progression. Am J Ophthalmol 2016;170:58–67.
36. Linsenmeier RA, Zhang HF. Retinal oxygen: from animals to humans. Prog Retin Eye Res 2017;58:115–151.
37. Yafai Y, Iandiev I, Lange J, et al. Müller glial cells inhibit proliferation of retinal endothelial cells via TGF-β2 and Smad signaling. Glia 2014;62:1476–1485.
38. Yafai Y, Lange J, Wiedemann P, et al. Pigment epithelium-derived factor acts as an opponent of growth-stimulatory factors in retinal glial-endothelial cell interactions. Glia 2007;55:642–651.
39. Yafai Y, Eichler W, Iandiev I, et al. Thrombospondin-1 is produced by retinal glial cells and inhibits the growth of vascular endothelial cells. Ophthalmic Res 2014;52:81–88.
Keywords:© 2019 by Ophthalmic Communications Society, Inc.
age-related macular degeneration; optical coherence tomography; type 3 neovascularization; volume rendering