Abstract
Purpose:
No large-mammal surgical models exist for geographic atrophy (GA), choroidal neovascularization (CNV), and pachychoroidal vascular remodeling. Our goal was to develop a porcine RPE debridement model of advanced macular degeneration to study photoreceptor cell loss and choroidal remodeling.
Methods:
Seven 2-month-old female domestic pigs were used for this study. After 25G vitrectomy, the area centralis was detached via subretinal bleb. A nitinol wire (Finesse Flex Loop) was used to debride RPE cells across a 3- to 5-mm diameter region. Fluid–air exchange was performed, and 20% SF6 gas injected. Animals underwent fundus photography, fluorescein angiography, optical coherence tomography (OCT), and OCT-angiography (OCTA) at 2 weeks, 1 month, 2 months, 3 months, and 6 months postoperatively. Retinal histology was obtained at euthanasia, 2 months (n = 3), 3 months (n = 2), or 6 months (n = 2) after surgery.
Results:
RPE debridement resulted in GA with rapid loss of choriocapillaris, progressive loss of photoreceptors, and pachychoroidal changes in Sattler's and Haller's layers in all seven eyes undergoing debridement within 2 months. OCT and histological findings included subretinal disciform scar with overlying outer retinal atrophy; outer retinal tubulations and subretinal hyper-reflective material. OCTA revealed type 2 CNV (n = 4) at the edges of the debridement zone by 2 months, but there was no significant exudation noted at any time point.
Conclusions:
Surgical debridement of the RPE results in GA, CNV, and pachychoroid and reproduced all forms of advanced macular degeneration. This surgical model may be useful in examining the role of RPE and other cell replacement in treating advanced macular disease.
AMD is the leading cause of irreversible vision loss among the elderly population worldwide.
1,2 Geographic atrophy (GA), the advanced form of nonexudative (dry) AMD, is characterized by the progressive degeneration of the RPE and the underlying choriocapillaris, leading to severe visual impairment owing to central scotoma.
3,4 Similarly, inherited macular dystrophies, such as Stargardt disease, central areolar choroidal dystrophy, and mitochondrial disorders (e.g., maternally inherited diabetes and deafness), among others, lead to profound central vision loss due to progressive macular degeneration in advanced stages. Stressors, genetic variants, chronic inflammation, and complement pathways have been implicated in macular degeneration pathobiology,
5–7 yet the temporal relationship of RPE loss, photoreceptor (PR) degeneration, and choroidal thinning remains elusive.
One approach to the study of GA has been the use of RPE debridement models. Although rodent,
8–11 rabbit,
12–14 and feline
15 models have been instrumental in elucidating certain aspects of macular degeneration pathogenesis, their anatomical and physiological differences from humans limit their translational relevance. Thus, many investigators have turned to either pigs or nonhuman primates. Across all mammalian eyes, the pig eye resembles the human eye second only to primates, both of which have holangiotic retinal vasculature.
16,17 Although a pig does not have a true macula, it does have a cone-rich central region of the retina that functions similarly to a human macula, allowing good daylight visual acuity and excellent color vision. Thus, the pig, like nonhuman primates, is a suitable model for macular and retinal degeneration research.
The first porcine RPE debridement model was introduced in the mid 1990’s by Del Priore et al.
18,19 This model used subretinal mitomycin C and edetic acid to debride the RPE, resulting in choriocapillaris loss after RPE loss. The model was hampered by RPE repopulation, which limited its use in further studies. Sodium iodate has been used to trigger a chemically induced RPE and PR degeneration in many species including swine, but does not target the RPE selectively.
20 Mechanical RPE debridement permits more selective targeting of the RPE and can be performed regionally in a large animal eye. Seah et al.
21 performed mechanical RPE debridement in a nonhuman primate model to examine the efficacy of an RPE cell transplantation; however, they did not study the effects of RPE removal on the adjacent outer retina and choroid. Similarly, in another study using nonhuman primates, confluent macular laser was shown to cause retina and choroidal damage; however, laser-induced damage makes it difficult to ascertain timing of RPE, PR, and choroidal changes.
22 Sharma et al.
23 used a similar laser model in swine to examine the efficacy of RPE transplantation, but like Seah et al.,
21 they did not study the effects of RPE removal on the adjacent outer retina and choroid.
This study presents the development of a porcine surgical model mimicking advanced macular degeneration using a mechanical RPE debridement technique. The retinal vascular structure of pigs closely mimic humans, making them an ideal choice for studying retinal diseases and regenerative therapeutic approaches.
18,20,24 RPE debridement allows investigation of the pathophysiological changes associated with macular atrophy, including the extent and timing of choroid and PR loss after selective RPE loss. A large animal RPE debridement model may be useful in examining the role of RPE and other cell replacement in treating advanced macular disease. This study also investigates the timing and pathophysiology of macular neovascular complexes (MNCs) in the setting of GA. MNCs, characterized by abnormal blood vessel growth in the macula, are a significant cause of vision loss in patients with AMD. However, the specific mechanisms and temporal relationship between MNCs and GA development remain poorly understood.
Last, the impact of RPE loss on choroidal flow and its subsequent effects on vessel density and thickness remain unknown. Understanding these dynamic changes in the choroid is crucial, because the choroid plays a vital role in maintaining retinal homeostasis and is implicated in the progression of GA. By using a debridement pig model with selective and permanent RPE removal, we can assess these parameters clinically using OCT and OCTA, providing further insights into the pathophysiology of GA and pachychoroid while complementing existing animal models.
Surgeries were performed essentially as before
25 with the following modifications. Before pars plana vitrectomy, indirect laser photocoagulation was performed 360° in the retinal periphery, anterior to the equator using a head-mounted 532 nM wavelength indirect laser ophthalmoscope (Purepoint, Alcon, Fort Worth, TX) to minimize the risk of post–pars plana vitrectomy retinal detachment. A 1- to 5-mm lateral canthotomy was performed on the right eye to improve exposure, and a 5-mL retrobulbar block of cefazolin (100 mg/mL) was administered into the subtenon's space superotemporally to proptose and stabilize the eye.
Three-port triamcinolone-assisted 25G pars plana vitrectomy using a Constellation Vitrectomy System (Alcon) and wide-angled, noncontact Biom fundus lens (Oculus Surgical, Wetzlar, Germany) was performed by one of two vitreoretinal surgeons (RI or BAS) with careful separation of the posterior hyaloid in the area centralis. Eye pressure was maintained using a balanced salt solution (BSS) infusion. A 38G polyimide subretinal cannula (38g PolyTip, MedOne, Sarasota, FL) was used to create a subretinal bleb using BSS with eye pressure decreased to 10 mm Hg. Injection of BSS was performed per surgeon preference, either with slow manual injection using a skilled assistant or with pneumatic foot-pedal control using the viscous fluid injector setting on the Constellation, with care to not exceed 12 pounds per square inch.
Endodiathermy was applied along the proximal bleb, and a 2-mm retinotomy was created with vertical scissors. A Finesse Flex Loop (Alcon) was used to gently debride and remove a 3- to 5-mm diameter region of RPE cells. Fluid–air exchange was performed, and 20% SF6 gas was injected. The sclerotomies were sutured with 8-0 Vicryl sutures, when indicated, to ensure airtight closure of all sclerotomy sites. The canthus was sutured with 4-0 chromic gut interrupted sutured. Gentamicin 0.3% drops were topically administered to the right eye three times daily for 5 days postoperatively.
Color Fundus Photography, Fluorescein Angiography, Optical Coherence Tomography (OCT), and OCT Angiography (OCTA)
Of the seven pig eyes described in this study, all developed common features in the areas of RPE debridement that mimicked several age-related and inherited macular diseases (
Fig. 7). First, all pigs developed large choroidal vessels in Sattler's and/or Haller's layers, a finding commonly observed in advanced macular degeneration. Ng et al.
29 found that more than one-half of the eyes (52.1%) with exudative maculopathy owing to AMD without a polypoidal phenotype had findings of pachychoroid, including increased choroidal thickness, choroidal vessels density, and/or vessel diameter (i.e., pachyvessels). These findings suggest that there are structural and functional alterations of the choroid (e.g., choroidal venous stasis, choroidal hyperpermeability) both in the human and the pig model as a direct result of RPE dysfunction. The presence of these vessels, especially in age-related GA, is especially intriguing because histological studies have shown that choroidal vessel size and density decrease with physiological aging.
29,30 Overall, the GA-associated pachyvessels and choriocapillaris thinning in this porcine model are consistent with histopathologic findings described in advanced AMD, polypoidal choroidal vasculopathy, and pachychoroid spectrum disorders.
29,31
Another common pig OCT finding that simulated human disease included presence of ORTs, which are common in areas of GA, CNV, and/or subretinal hyper-reflective material.
32 Hypertransmission defects in areas of cRORA, subretinal hyper-reflective material, pigment migration, and disciform scar formation were also found within the debrided regions and were nearly indistinguishable from the human examples of macular degeneration. The collapsed outer retinal layers noted in several pigs, most prominently seen in
Figure 4 at the transition zone between GA and healthy retina, seem to be consistent with the external limiting membrane descent described by Li et al.
33 Histological staining supports the hypothesis that reactive gliosis and disciform scar formations may be due to microglia and astrocyte recruitment to the area of RPE loss.
Before this study, other debridement animal models have been described using various species and both mechanical and chemical debridement techniques. These models have advanced our understanding of GA in the context of macular degeneration and have been effectively used to optimize subretinal cell transplantation. However, models such as that described by Monés et al.
20 using sodium iodate are not specific to RPE cells. The use of mitomycin C and edetic acid for chemically induced GA was described in the pig in 1995, before the use of OCTA.
18,19 Although that study clearly demonstrated with OCT and immunocytochemistry that RPE loss led to choriocapillaris atrophy, RPE repopulation in the debridement zone was a consistent confounding factor and a phenomenon we hoped to mitigate with mechanical debridement. Although both mechanical and laser based methods of debridement have been applied to nonhuman primates and pigs, these studies sought to characterize RPE transplantation immediately following debridement and did not perform a comprehensive analysis of the debridement region absent transplanted RPE.
21,23 Our study is the first to do so.
It is important to note that our model shows that RPE loss affects choroidal flow, because a complete flow deficit was noted as early as 2 weeks after RPE debridement. None of the seven pigs had significant subretinal hemorrhage at the time of debridement, suggesting that the consistent and progressive loss of choroid was not due to surgical trauma. One would expect massive bleeding, especially in a pig, if there was significant trauma to the choroid at the time of RPE removal. This finding corroborates clinical OCTA evidence of severe impairment of choriocapillaris flow within the bounds of GA.
34 However, it is traditionally taught that choroidal circulation changes lead to overlying RPE dysfunction, and eventually death, in advanced AMD and other macular diseases.
5,29 Our study shows that the opposite can also occur, where loss of RPE negatively impacts the choroid. These data support the findings of McLeod, et al.
35 In that study, post mortem analysis of eyes with advanced AMD showed a 50% decrease in choroidal vascular area and capillary constriction in regions of complete RPE atrophy.
35 Their group concluded that the primary insult in GA is at the level of the RPE, which is consistent with our findings. RPE cells are known to secrete competing pro and antiangiogenic factors (e.g., pigment epithelium-derived factor VEGF. Thus, one hypothesis for RPE loss leading to choroidal atrophy is the potential imbalance between proangiogenic and antiangiogenic factors in the debridement zone.
This porcine model also shows that the choroid atrophies faster than PRs, suggesting that PR loss is, at least partly, owing to choroidal atrophy (i.e., ischemia) and not solely a lack of outer segment phagocytosis. In the work of Li et al.,
33 histological analysis of GA confirmed end-stage loss of PRs and surrounding supporting structures and vertically oriented cells (e.g., Müller cells). Their study highlights the irreversible tissue damage and loss after RPE loss. Given that our model preserves outer retinal structures initially, it is feasible that immediate replacement of RPE cell sheets could potentially restore structure and function in the debridement area in this pig model. Our laboratories have ongoing translational studies to examine the role of RPE and other cell replacement in this porcine debridement model.
Retinal damage, especially the loss of RPE, has been associated with increased injection pressure during subretinal gene therapy delivery in various animal models, including pigs.
24 Unfortunately, significant perifoveal chorioretinal atrophy after subretinal gene therapy injection is now recognized as a common complication in human surgeries of voretigeneneparvovec-rzyl (Luxturna), a U.S. Food and Drug Administration–approved gene therapy for patients with biallelic RPE65 variants.
36,37 This chorioretinal atrophy in human gene therapy surgeries is not well-understood. Hypotheses include the shearing of RPE owing to increased injection pressures (i.e., significant flow) during bleb propagation and/or viral solution toxicity to the RPE, both of which are plausible explanations. In these cases, which are associated with paracentral scotomas, the RPE, outer retinal layers, and inner choriocapillaris are all thinned or missing, similar to the porcine model we describe here.
Our data, which showed RPE loss first followed by choroid thinning and then PR loss, corroborates the histological findings of Del Priore et al.
18,19 Together these studies suggest that the RPE damage from subretinal bleb propagation (from shearing forces) and/or viral toxicity could lead to secondary inner choroid thinning, subsequent outer retinal loss, and eventually choroidal atrophy. Given the junctional, or transitional, zone between the debridement region and nondebrided regions showed significant pathology, including choroidal thinning, pachyvessels, and CNV formation, it is reasonable to believe the choroid-related changes in the junctional zone caused by RPE injury may be responsible for the progressive nature of the perifoveal chorioretinal atrophic regions in human subretinal gene therapy.
Another unsolved and controversial topic in the field of retina is the potential association between anti-VEGF agents and the progression of GA.
38 Although anti-VEGF therapies have revolutionized the treatment of neovascular (wet) AMD, concerns have been raised regarding their impact on the development or progression of GA. One study showed that central choroidal thickness decreased significantly after anti-VEGF therapy for neovascular AMD, possibly implicating choroidal ischemia and subsequent overlying RPE dysfunction in GA development and/or progression.
39 Understanding these potential associations necessitates a deeper understanding of the pathophysiology of GA, and future studies exploring this association using this pig model could elucidate the underlying mechanisms of GA expansion.
Because this RPE debridement model results in CNV formation in both the GA and disciform scar phenotypes, future preclinical anti-VEGF studies could be designed using this pig model to test efficacy against CNV formation, exudation, and expansion. It is possible that the nitinol wire used for debridement led to small breaks in Bruch's membrane without resulting in clinically significant hemorrhage at the time of surgery. These small microperforations could eventually lead to CNV development. However, all observed CNVs developed at the junctional zones, which is often observed in patients with GA-associated CNVs. We hypothesize that the areas devoid of RPE underwent choroidal remodeling with loss of choriocapillaris, localized choroidal ischemia, and subsequent CNV formation. It is possible that the alterations in choriocapillaris perfusion and the absence of an effective blood-retina barrier allow leakage of pro-angiogenic factors from RPE at the edges of the debridement zone to the choroid and promote blood vessel growth from the healthy choroid at its edges. Regardless of the etiology, the presence of CNV in this model was both unexpected and welcomed by our team for future research endeavors.
The authors thank Evan L. Atherton for his contributions in preparing the operating room and instrumentation for the pig surgeries. The authors also thank the Mayo Clinic Veterinary staff and the talented veterinary technicians in the Department of Surgery at Mayo clinic, especially Joanne Pedersen, Lisa Yngsdal, and Andrew Alden.
The authors acknowledge the support of the Mayo Clinic Foundation for Biomedical Research, the Mayo Clinic Department of Ophthalmology, and Seeing Medicines Inc.
Meeting Presentation: Platform presentation at the Association for Research in Vision and Ophthalmology (ARVO) Annual Meeting May 2022, Denver, Colorado.
Disclosure: R Iezzi, Jansen (C); B.A. Scruggs, Genentech and Iveric Bio (C); J. Gandhi, None; F.N. Zenti, None; N. Shafi, None; A. Berger, None. A.D. Marmorstein, None