- Email: [email protected]
Analysis of animal models of macular edema
SURVEY OF OPHTHALMOLOGY
VOLUME 28. SUPPLEMENT * MAY 1984
Analysis of Animal Models of Macular Edema ROY W. BELLHORN,
D.V.M.
Department of Ophthalmology, Montejiore Medical Center, Albert Einstein College of Medicine, Bronx, New York
Abstract. Various
models of macular edema have been studied; however, frank development of a prototypical cystoid macular edema has not been observed. In humans, cystoid macular edema is frequently observed in association with other disturbances of the retina. Thus, a basic drawback of the animal models may be that an otherwise healthy retina is capable of resolving the experimentally produced edema, thereby preventing chronic cystoid maculopathy. A review of macular edema models and of experimental retinal and brain edema investigations suggests that blood-retinal (blood-brain) barrier permeability abnormalities need to be accompanied by ineffective edema resolving mechanisms for the production of a chronic edema. Intraglial uptake of extravasated serum proteins has been hypothesized to be an edema-resolving mechanism in brain edema. As such, the hypothesis that the Miiller cell may be important to edema resolution appears attractive. Future animal model studies should include methodologies whereby edema resolution mechanisms are impaired. (Surv Qphthalmol 28(Suppl):520-524, 1984)
Key words. macular
edema
animal l
models l blood-ocular barriers retinal anatomy l retinal edema
C
l l
cystoid macular edema subhuman primates
l
sands of subhuman primates over the past 20 years encountered cystoid macular edema. It is possible that this entity occurs in subhuman primates, but it is understandably difficult to Jocate the affected animals since a subjective complaint of visual impairment - the factor that brings most human cases to the ophthalmologist - is not communicated to us. It is obvious, therefore, that the development of appropriate experimental animal models provides the best opportunity for investigating cystoid macular edema. Dr. Mark Tso has been the most active investigator concerned with experimental macular edema. I have analyzed sbme of his models in an attempt to provide some insight into ways of enhancing the development of future models.
ystoid macular edema occurs in conjunction with a variety of ocular conditions. Fortunately, in many instances the edema is transient and its resolution allows a return of visual acuity to preedematous levels. However, if edema resolution is prolonged, various degrees of visual loss may result. Because of the potentially serious effects of macular edema, ophthalmic researchers should acquire an understanding of the pathophysiological mechanisms responsible for the development and the resolution or nonresolution of retinal edema. Animal models provide the researcher an opportunity to investigate disease mechanisms at any stage of a pathological process, and also to utilize cellular and subcellular investigative techniques not applicable to clinical research. In general, optimal models are those in which the disease process occurs spontaneously, and those that are predictably reproducible. Since macular edema involves a certain specialized portion of the retina, our optimal model would be its spontaneous occurrence in a foveate primates. Reviewing species, e.g., the subhuman the literature, I have not found a single report of spontaneous cystoid macular edema in subhuman primates. Nor have I in my examination of thou-
Photic Maculopathy Model In 1972, Tso et alI7 reported that visible light damage to the fovea of rhesus monkeys caused macular edema and that this edema subsided after one week. Light microscopy showed the edema to involve all retinal layers. A subsequent electron microscopic examination of tissues of up to five months postexposure revealed no other evidence of permeability abnormalities. I4 Tso later reported, however, 520
ANIMAL
MODELS
OF MACULAR
521
EDEMA
that a monkey at four years postexposure evidenced retinal pigment epithelial permeability abnormalities to both sodium fluorescein and horseradish peroxidase (HRP), and that histologically there was mild edema of the outer plexiform layer.12 Tso also noted that in spite of the chronic leakiness of the retinal pigment epithelial barrier, there was no evidence df photoreceptor damage.‘* Comment. This model demonstrates that a chronic permeability abnormality of the retinal pigment epithelium (RPE) to horseradish peroxidase, a protein about the size of serum albumin, led only to a mild edema with no evidence of photoreceptor damage (but would there be a demonstrable loss of visual acuity?). The importance of this model is that although proteins chronically passed an altered RPE barrier, only mild intraretinal edema ensued. Thus, this model possibly represents an opportunity to study mechanisms of edema resolution, and these mechanisms are an important aspect of our search to understand cystoid macular edema.
Ocular Hypotony Model Ciliary body cryotherapy in rhesus monkeys by Tso and Shih caused ocular hypotony with subsequent disruption of the blood-retinal barriers with some emphasis in the macular region.ls Electron microscopy demonstrated the passage of HRP across both the retinal vessel and pigment epithelial barriers in two monkeys; in two others, HRP only passed the retinal vessel barrier. However, evidence of retinal edema was not reported, but there was evidence of HRP in sdme Miiller cells. Comment. Passage of proteins across the bloodretinal barrier over a period of five weeks did not produce retinal kdema in this model. It is possible that a longer period of time would have led to edema development; however, Tso did examine one eye at 16 weeks and found that no permeability abnormality existed at that time. In essence, the barriers had undergone repair. This model would provide a basis for studying mechanisms whereby acute retinal edema formation, in spite of leaky barriers, is prevented. One very interesting observation was the presence of HRP within some Miller cells.
Aphakic Model Cystoid macular edema following cataract extraction in man (Irvine-Gass-Norton syndrome) is a commonly observed phenomenon. Tso and Shih” performed intra- and extracapsular lens extractions with and without vitreous loss in rhesus monkeys in order to study this syndrome. Electron microscopic examination of tissues up to 30 days postoperatively demonstrated the passage of HRP into the neural retina across both the retinal vessel and pigment
epithelial barriers in those monkeys that had experienced vitreous loss. Passage of HRP across just the RPE barrier was evident in eyes without vitreous loss. In their initial report, the authors noted that cystoid accumulation of fluid in Henle’s layer was not seen. In a later report, Tso noted that swollen Miiller cells with a watery cytoplasm were observed in the aphakic model.‘“, Comment. Tso and Shih suggested that lack ofedema formation in this model may have been the result of using young healthy monkeys, i.e., a healthy retina is not conducive to development of cystoid macula edema. I concur with their thought, but also wonder if, again, a longer follow-up period may have shown different results. Irvine has reported that while some cystoid macular edema can be observed occasionally at four weeks postoperatively, it is most likely to be seen at 6-8 weeks.” I also wonder if the swollen, watery Miller cells should perhaps be considered to represent intracellular edema? Fine and Brucker found swollen Miiller cells to be a basic pathologic process in their ultrastructural study of human cystoid macular edema.” Possibly in this animal model Tso and Shih did demonstrate the early beginnings of a cystoid macular edema and, thus, it could be a very important one to pursue. This model also appears to be an excellent one for studying the edema resolution mechanisms within normal retinas that prevent development of cystic edema.
Argon Laser Fovea1 Damage Model Following damage to the fovea of rhesus monkeys by argon laser, Tso and Fine examined the tissues by electron microscopy over a four-year period.‘” At the three-year investigation, extracellular vacuolation of the RPE layer was evident and there was mild extracellular vacuolation of Henle’s layer. At the four-year study, Henle’s layer showed the presence of extracellular cystoid spaces. Comment. Unfortunately, no permeability tracer studies were performed on this model and, therefore, we do not know the status of the barrier tissues. It is interesting that the subsequent edema formation in this model was extracellular, in contrast to the previously mentioned intracellular (Miller) edema observed in the three human cases.? This model points out (as have several others) the importance of longterm studies, again recognizing the ability of otherwise healthy retinal tissues to resist edema development.
Talc Retinopathy Model The final model for consideration is that of macular edema associated with experimental talc retinopathy in rhesus monkeys, which was reported recently by Jampol” and Saga et a1.7.8 Repeated
522
Surv Ophthalmol
28(Suppl)
May 1984
intravenous injections of talc over a period of 3?&10 months resulted in multiple microinfarcts of the retinal and choroidal arterioles and capillaries, especially in the macula. Permeability abnormalities to sodium fluorescein and to HRP were evidenced by the retinal vessels but not the pigment epithelium. Ultrastructural abnormalities included extracellular cystoid spaces within the inner retinal layers and the outer plexiform layer. The authors note that this model centers on chronic microinfarction of the retinal vessels, including those in the macula, and that this could be an important pathogenetic mechanism in cystoid macular edema. Fine and Brucker” had earlier postulated a similar mechanism, having found evidence of microinfarction in their ultrastructural study of human cystoid macular edema; they thought that focal ischemia could have led to Miiller cell abnormalities. Comment. This model is most interesting and quite informative. It suggests that chronic vascular insult is important and that the insult may center not only on permeability abnormalities of the blood-retinal barriers, but also on ischemia. In spite of the fact that this model demonstrated extracellular edema in rhesus monkeys, whereas Fine and Brucker demonstrated intracellular edema in humans, this model may demonstrate that more than one pathogenetic mechanism is involved in macular edema. Tso’s recent report on the pathology of macular edema in humans”’ indicates that the edema can be present in different retinal layers in different clinical ocular diseases. Thus it is not suprising to encounter divergent findings in different experimental circumstances. The possibility also exists that talc retinopathy over a longer period of time could lead to an intracellular (Miiller) edema. In any event, this model provides a firm basis for further studies.
Discussion At a Cambridge Ophthalmic Symposium several years ago, Tso commented upon his studies of macular edema in animal models: “These studies suggest that it is incorrect to assume that petaloid staining of microcysts in cystoid macular edema is a simple accumulation of leaked fluorescein from disruption of the blood-retinal barrier. Pathogenetic mechanisms other than simple disruption of the blood-retinal barriers may be responsible for the formation of microcysts of the macula.“” I wholeheartedly concur with this statement and would like to suggest that Miiller cell dysfunction is one of those pathogenetic mechanisms. Fine and Brucker stated that “the edematous (macular) process apparently begins with intracytoplasmic swelling of Miiller cells, itself probably secondary to the vascular abnormahties.“” The question is, what causes
BELLHORN the Miiller cell to swell? Fine and Brucker answer to that question as follows: “The mechanism of or reason for the edema of the Miiller cell is not clear. It may be the result of a faulty retinal vasculature, a faulty or incompetent retinal pigment epithelium, by some as yet inadequately understood mechanism, or some combination thereof.” I believe that Tso’s work with animal models indicates that “or some combination thereof’ is most likely. The human macula is but one portion of the retina, and it does exhibit an exaggerated edematous response to a widespread retinal disturbance. One may ask whether information concerning retinal edema in general can help us understand macular edema? In 1979, Drs. Henkind, Schall and I reviewed the pertinent literature concerning retina edema.+ We found that while much had been written about the clinical aspects of human retinal edema, little attention had been paid to the mechanism(s) of development and course. Some relevant observations were made by Kohner et allo in their study of branch retinal vein occlusion in rhesus monkeys. They noted that the Miiller cell was an integral factor in the edematous process, and in their discussion they stated: “The distribution of The anaretinal edema . . . is not fully understood. tomic arrangement and functional differences within the Miiller cell in different retinal layers may each play a part. . .” They further state that “perhaps in the inner nuclear layer the Miller cell takes up fluid actively to preserve the surrounding neural elereinforce the thoughts of ments.” These statements Fine and Brucker in connection with their finding of macular microinfarction: “Such vascular changes, of course, can produce areas of retinal ischemia, to which the Miiller cells in this region of the foveomacula would be vulnerable. This vulnerability may be related to the morphologic observation that foveomacular Miller cells occupy a greater volume and have a more uniform lucency of cytoplasm than they do elsewhere in the retina.“” In contrast to the paucity of investigations dealing specifically with retinal edema, experimental brain edema has been extensively studied, probably because of the life threatening circumstances associated clinically with brain edema. A recent volume of Advances in Neurology (1980) entitled Brain Edema provides considerably background for investigating and understanding retinal edema. What the neuroscientists term “vasogenic brain edema” is of particular interest to us, and in the volume noted above, Klatzo et al” authored a most pertinent article on the resolution of vasogenic brain edema. Based on a series of experiments, they proposed the following hypothesis concerning development and resolution of brain edema. In vasogenic brain ede-
ANIMAL
MODELS OF MACULAR
523
EDEMA
ma, injured vessels allow edema fluid and serum proteins to enter the extracellular space. Astrocytes are swollen by edema fluid and proteins; the subsequent intracellular lysomal digestion of the proteins then provides the basic mechanism for removing proteins from the extracellular space. This protein removal allows a return to normal of colloidal osmotic forces which in turn allows water to move out of the extracellular space across the blood-brain barrier.” Thus, edema resolution is hypothesized to be dependent upon removal of osmotically active proteins by intraglial uptake of the extravasated serum proteins. Since the principal glial cell of the retina is the Miiller cell, we may also hypothesize that prevention and resolution of retinal edema is dependent upon Miiller cell function (with, of course, appropriate return to normal of the leaky blood-retinal barriers). Thus, if blood-retinal barrier function is reestablished and Miiller cell function remains operative, macular edema may not occur. Tso’s ocular hypotony model may represent this in that (1) leakage without edema occurred for a short period of time, (2) HRP was observed in Miiller cells, and (3) shortly afterward the blood-retinal barriers no longer leaked proteins. In essence, this is an edema prevention model. Retinal edema was also not observed in the aphakit edema model even though leakage of proteins was demonstrated during the 30-day postoperative period. But again the edema resolution mechanisms (the Miiller cell?) were apparently capable of preventing the development of macular edema. In the photic maculopathy model the acute edematous process was resolved, but over a period of almost four years, a persistent leak ofserum proteins may have begun to tax the Miiller cell such that mild edema of the outer plexiform (Henle’s) layer became evident. We do not know if that edema was extracellular, intracellular or if it involved the Miiller cell. But, as we can now see, these questions are important. On the other hand, after a period of three to four years, the argon laser maculopathy model showed vacuolation (edema?) of the extracellular space in Henle’s layer. There was also evidence of proteinaceous material in the enlarged (cystoid?) extracellular spaces. This model could therefore represent a mild but chronic leakage of proteins across the blood-retinal barriers and an impaired ability of the Miiller cell to take in the extravasated proteins. It is possible, for example, that the laser had destroyed a significant portion of the macular component of Miiller cells. Talc retinopathy with macular edema, as noted earlier, is both interesting and informative. In that
situation we have a multiple and prolonged insult to the blood-retinal barriers, producing sufficient impact upon the neural retina to cause a number of retinal layers to develop cystoid edema. It would seem that the extracellular cystic spaces are in contrast to the intracellular (Miiller) cystoid edema described in the human cases, but even so, this model presents an important opportunity to study development and nonresolution of macular edema. We can now see that Tso’s statement in his Cambridge Symposium paper was indeed prophetic, that indeed pathogenetic mechanisms other than simple disruption of the blood-retinal barriers are necessary for the development of cystoid macular edema.” I sue;qest that Miiller cell dysfunction may IL he one of those mechanisms. There certainly are other mechanisms responsible for movement of fluid and other substances through the retinal tissue (e.g., the Na+ K’ ATPase pump) and undoubtedly it is necessary to investigate these biochemical mechanisms as well. At the same Cambridge Symposium, I reported on work in our laboratory and by other investigators that indicated the importance of the retinal environment upon the morphologic characteristics of developing and/or existing retinal vessels.’ It would appear, based upon our present concepts that we may need to manipulate the retinal environment as well as disrupt the blood-retinal barriers in order to develop models of chronic cystoid macular edema. That is, we need to create an environment that is not conducive to the resolution of the edematous process. There are various ways that this could be accomplished. For example, permeability abnormalities in the blood-retinal barrier have been demonstrated in monkeys with experimental hypertension” or diabetes.” If there were a macular insult superimposed on these abnormal retinas, would a cystoid macular edema result? .4nother approach would he to use selected cytotoxic agents to impair edema resolution mechanisms. For example, sodium iodate damages the integrity of the retinal pigment epithelial cells, oubain interferes with Na+ K+ XTPase pump mechanisms, alpha amino-adipic acid can damage the Miller cell, etc. Can we, through the use of agents such as these, selectively produce various types of retinal edema? I believe that we must if we are to develop suitable animal models which will enable us to understand the mechanisms responsible for the development, resolution and nonresolution of macular edema.
References
524
Surv Ophthalmol
28(Suppl)
May
1984
mol 6% Sci 19(ARVO .Suppl):4, 1980 3. Fine BS, Brucker A.J: Macular edema and cystoid macular edema. AmJ Ophthalmol 92:466-481, 1981 4. Henkind P, Bellhorn RW, Schall B: Retinal edema: postulated mechanism(s), in Cunha-Vaz.JG (ed): The Blood-Retinal Barriers. New York, Plenum Publishing, 1980, pp 251-268 5. Irvine AR: Cystoid maculopathy. Sure Ophthalmol21: I-1 7, 1976 6. ,Jampol LM. Setogawa T, Rednam KRV. Tso MOM: Talc retinopathy in primates. A model of ischemic retinopathy: I. Clinical studies. Arch Ophthalmol 99: 1273-1280, 1981 7. Kaga N, Tso MOM. Jampol LM, et al: Talc retinopathy in primates: A model of ischemic retinopathy. II. A histopathologic study. Arch Ophthalmol 1011:1644-1648, 1982 8. Kaga N. Tso .MOM,Jampol LM: Talc retinopathy in primates: A model of ischemic retinopathy. 111. An electron microscopic study. Arch Ophthalmol 100: 1649-1657, 1982 9. Klatzo I, Chui E, Fu,jiwara K, Spatz M: Resolution ofvasogenic brain edema, in Cervos-Xavarro .J. Fcrszt R (cds): .4dza:a,ws in .Ye’eurologv /[‘al 28) Brain Edema. NW York. Raven Press, 1980, pp 359-373 IO. Kohner EM, Dollcry CT. Shakib Xl. ct al: Experimental retinal branch vein occlusion. rim J Ophthaltncl 6Y:778-825, 1970 11. Tso MOM: Pathological study of cystoid macular ordema. Tram Ophthalmol Sac (‘I( 100:408-413, 1980 12. Tso X1011: Pathology of the blood-retinal barrier, in CunhaVaz,JG (rd): T/w Blood-Rebnal Barriers. New York, Plenum Puhlishing, 1980, pp 235-250
BELLHORN 13. ‘l’s0 MOM: Pathology of cystoid macular edema. Ophthalmology R9:902-9 14, I982 14. Tso MOM: Photic maculopathy in rhesus monkey. A light and electron microscopic study. InvestOphthalmol 22: 17-34, 1973 15. Tso MOM: Cunha-Vaz J, Blair N, et al: Disruption of bloodretinal harrier in diabetic primates. Invest Ophthalmol Vis Sci -32/8RVO Suppl): 110, 1982 16. Tso MOM, Fine BS: Repair and late degeneration of the primate foveola after injury by argon laser. InueslOplrlhalmof Vis Sri l&447-461, 1979 17. ‘l‘s0 MOM, Fine BS, Z’lmmerman LE: Photic maculopathy produced by the indirect ophthalmoscope. I. Clinical and histopathologic study. Am J Ophthalmoi 73:68&699, 1972 18. Tso MOM. Shih CY: Disruption of blood-retinal harrier in ocular hypotony: Preliminary report. Exp $ve Res 23:20%216, 1976 19. ‘I’s0 MOM, Shih CY: Experimental macular edema after lens extraction. lnmt Ophthalmof Vis Sri 16:381-392, 1977
Supported by NE1 Grant EY 02038 and hy Research to Prevent Blindness, Inc. Reprint requests should be addressed to Roy Bellhorn, D.V.M., hlhrrt Einstein College of Medicine, I I1 East 210th Street, Bronx, NY 10467.
VOLUME 28. SUPPLEMENT * MAY 1984
Analysis of Animal Models of Macular Edema ROY W. BELLHORN,
D.V.M.
Department of Ophthalmology, Montejiore Medical Center, Albert Einstein College of Medicine, Bronx, New York
Abstract. Various
models of macular edema have been studied; however, frank development of a prototypical cystoid macular edema has not been observed. In humans, cystoid macular edema is frequently observed in association with other disturbances of the retina. Thus, a basic drawback of the animal models may be that an otherwise healthy retina is capable of resolving the experimentally produced edema, thereby preventing chronic cystoid maculopathy. A review of macular edema models and of experimental retinal and brain edema investigations suggests that blood-retinal (blood-brain) barrier permeability abnormalities need to be accompanied by ineffective edema resolving mechanisms for the production of a chronic edema. Intraglial uptake of extravasated serum proteins has been hypothesized to be an edema-resolving mechanism in brain edema. As such, the hypothesis that the Miiller cell may be important to edema resolution appears attractive. Future animal model studies should include methodologies whereby edema resolution mechanisms are impaired. (Surv Qphthalmol 28(Suppl):520-524, 1984)
Key words. macular
edema
animal l
models l blood-ocular barriers retinal anatomy l retinal edema
C
l l
cystoid macular edema subhuman primates
l
sands of subhuman primates over the past 20 years encountered cystoid macular edema. It is possible that this entity occurs in subhuman primates, but it is understandably difficult to Jocate the affected animals since a subjective complaint of visual impairment - the factor that brings most human cases to the ophthalmologist - is not communicated to us. It is obvious, therefore, that the development of appropriate experimental animal models provides the best opportunity for investigating cystoid macular edema. Dr. Mark Tso has been the most active investigator concerned with experimental macular edema. I have analyzed sbme of his models in an attempt to provide some insight into ways of enhancing the development of future models.
ystoid macular edema occurs in conjunction with a variety of ocular conditions. Fortunately, in many instances the edema is transient and its resolution allows a return of visual acuity to preedematous levels. However, if edema resolution is prolonged, various degrees of visual loss may result. Because of the potentially serious effects of macular edema, ophthalmic researchers should acquire an understanding of the pathophysiological mechanisms responsible for the development and the resolution or nonresolution of retinal edema. Animal models provide the researcher an opportunity to investigate disease mechanisms at any stage of a pathological process, and also to utilize cellular and subcellular investigative techniques not applicable to clinical research. In general, optimal models are those in which the disease process occurs spontaneously, and those that are predictably reproducible. Since macular edema involves a certain specialized portion of the retina, our optimal model would be its spontaneous occurrence in a foveate primates. Reviewing species, e.g., the subhuman the literature, I have not found a single report of spontaneous cystoid macular edema in subhuman primates. Nor have I in my examination of thou-
Photic Maculopathy Model In 1972, Tso et alI7 reported that visible light damage to the fovea of rhesus monkeys caused macular edema and that this edema subsided after one week. Light microscopy showed the edema to involve all retinal layers. A subsequent electron microscopic examination of tissues of up to five months postexposure revealed no other evidence of permeability abnormalities. I4 Tso later reported, however, 520
ANIMAL
MODELS
OF MACULAR
521
EDEMA
that a monkey at four years postexposure evidenced retinal pigment epithelial permeability abnormalities to both sodium fluorescein and horseradish peroxidase (HRP), and that histologically there was mild edema of the outer plexiform layer.12 Tso also noted that in spite of the chronic leakiness of the retinal pigment epithelial barrier, there was no evidence df photoreceptor damage.‘* Comment. This model demonstrates that a chronic permeability abnormality of the retinal pigment epithelium (RPE) to horseradish peroxidase, a protein about the size of serum albumin, led only to a mild edema with no evidence of photoreceptor damage (but would there be a demonstrable loss of visual acuity?). The importance of this model is that although proteins chronically passed an altered RPE barrier, only mild intraretinal edema ensued. Thus, this model possibly represents an opportunity to study mechanisms of edema resolution, and these mechanisms are an important aspect of our search to understand cystoid macular edema.
Ocular Hypotony Model Ciliary body cryotherapy in rhesus monkeys by Tso and Shih caused ocular hypotony with subsequent disruption of the blood-retinal barriers with some emphasis in the macular region.ls Electron microscopy demonstrated the passage of HRP across both the retinal vessel and pigment epithelial barriers in two monkeys; in two others, HRP only passed the retinal vessel barrier. However, evidence of retinal edema was not reported, but there was evidence of HRP in sdme Miiller cells. Comment. Passage of proteins across the bloodretinal barrier over a period of five weeks did not produce retinal kdema in this model. It is possible that a longer period of time would have led to edema development; however, Tso did examine one eye at 16 weeks and found that no permeability abnormality existed at that time. In essence, the barriers had undergone repair. This model would provide a basis for studying mechanisms whereby acute retinal edema formation, in spite of leaky barriers, is prevented. One very interesting observation was the presence of HRP within some Miller cells.
Aphakic Model Cystoid macular edema following cataract extraction in man (Irvine-Gass-Norton syndrome) is a commonly observed phenomenon. Tso and Shih” performed intra- and extracapsular lens extractions with and without vitreous loss in rhesus monkeys in order to study this syndrome. Electron microscopic examination of tissues up to 30 days postoperatively demonstrated the passage of HRP into the neural retina across both the retinal vessel and pigment
epithelial barriers in those monkeys that had experienced vitreous loss. Passage of HRP across just the RPE barrier was evident in eyes without vitreous loss. In their initial report, the authors noted that cystoid accumulation of fluid in Henle’s layer was not seen. In a later report, Tso noted that swollen Miiller cells with a watery cytoplasm were observed in the aphakic model.‘“, Comment. Tso and Shih suggested that lack ofedema formation in this model may have been the result of using young healthy monkeys, i.e., a healthy retina is not conducive to development of cystoid macula edema. I concur with their thought, but also wonder if, again, a longer follow-up period may have shown different results. Irvine has reported that while some cystoid macular edema can be observed occasionally at four weeks postoperatively, it is most likely to be seen at 6-8 weeks.” I also wonder if the swollen, watery Miller cells should perhaps be considered to represent intracellular edema? Fine and Brucker found swollen Miiller cells to be a basic pathologic process in their ultrastructural study of human cystoid macular edema.” Possibly in this animal model Tso and Shih did demonstrate the early beginnings of a cystoid macular edema and, thus, it could be a very important one to pursue. This model also appears to be an excellent one for studying the edema resolution mechanisms within normal retinas that prevent development of cystic edema.
Argon Laser Fovea1 Damage Model Following damage to the fovea of rhesus monkeys by argon laser, Tso and Fine examined the tissues by electron microscopy over a four-year period.‘” At the three-year investigation, extracellular vacuolation of the RPE layer was evident and there was mild extracellular vacuolation of Henle’s layer. At the four-year study, Henle’s layer showed the presence of extracellular cystoid spaces. Comment. Unfortunately, no permeability tracer studies were performed on this model and, therefore, we do not know the status of the barrier tissues. It is interesting that the subsequent edema formation in this model was extracellular, in contrast to the previously mentioned intracellular (Miller) edema observed in the three human cases.? This model points out (as have several others) the importance of longterm studies, again recognizing the ability of otherwise healthy retinal tissues to resist edema development.
Talc Retinopathy Model The final model for consideration is that of macular edema associated with experimental talc retinopathy in rhesus monkeys, which was reported recently by Jampol” and Saga et a1.7.8 Repeated
522
Surv Ophthalmol
28(Suppl)
May 1984
intravenous injections of talc over a period of 3?&10 months resulted in multiple microinfarcts of the retinal and choroidal arterioles and capillaries, especially in the macula. Permeability abnormalities to sodium fluorescein and to HRP were evidenced by the retinal vessels but not the pigment epithelium. Ultrastructural abnormalities included extracellular cystoid spaces within the inner retinal layers and the outer plexiform layer. The authors note that this model centers on chronic microinfarction of the retinal vessels, including those in the macula, and that this could be an important pathogenetic mechanism in cystoid macular edema. Fine and Brucker” had earlier postulated a similar mechanism, having found evidence of microinfarction in their ultrastructural study of human cystoid macular edema; they thought that focal ischemia could have led to Miiller cell abnormalities. Comment. This model is most interesting and quite informative. It suggests that chronic vascular insult is important and that the insult may center not only on permeability abnormalities of the blood-retinal barriers, but also on ischemia. In spite of the fact that this model demonstrated extracellular edema in rhesus monkeys, whereas Fine and Brucker demonstrated intracellular edema in humans, this model may demonstrate that more than one pathogenetic mechanism is involved in macular edema. Tso’s recent report on the pathology of macular edema in humans”’ indicates that the edema can be present in different retinal layers in different clinical ocular diseases. Thus it is not suprising to encounter divergent findings in different experimental circumstances. The possibility also exists that talc retinopathy over a longer period of time could lead to an intracellular (Miiller) edema. In any event, this model provides a firm basis for further studies.
Discussion At a Cambridge Ophthalmic Symposium several years ago, Tso commented upon his studies of macular edema in animal models: “These studies suggest that it is incorrect to assume that petaloid staining of microcysts in cystoid macular edema is a simple accumulation of leaked fluorescein from disruption of the blood-retinal barrier. Pathogenetic mechanisms other than simple disruption of the blood-retinal barriers may be responsible for the formation of microcysts of the macula.“” I wholeheartedly concur with this statement and would like to suggest that Miiller cell dysfunction is one of those pathogenetic mechanisms. Fine and Brucker stated that “the edematous (macular) process apparently begins with intracytoplasmic swelling of Miiller cells, itself probably secondary to the vascular abnormahties.“” The question is, what causes
BELLHORN the Miiller cell to swell? Fine and Brucker answer to that question as follows: “The mechanism of or reason for the edema of the Miiller cell is not clear. It may be the result of a faulty retinal vasculature, a faulty or incompetent retinal pigment epithelium, by some as yet inadequately understood mechanism, or some combination thereof.” I believe that Tso’s work with animal models indicates that “or some combination thereof’ is most likely. The human macula is but one portion of the retina, and it does exhibit an exaggerated edematous response to a widespread retinal disturbance. One may ask whether information concerning retinal edema in general can help us understand macular edema? In 1979, Drs. Henkind, Schall and I reviewed the pertinent literature concerning retina edema.+ We found that while much had been written about the clinical aspects of human retinal edema, little attention had been paid to the mechanism(s) of development and course. Some relevant observations were made by Kohner et allo in their study of branch retinal vein occlusion in rhesus monkeys. They noted that the Miiller cell was an integral factor in the edematous process, and in their discussion they stated: “The distribution of The anaretinal edema . . . is not fully understood. tomic arrangement and functional differences within the Miiller cell in different retinal layers may each play a part. . .” They further state that “perhaps in the inner nuclear layer the Miller cell takes up fluid actively to preserve the surrounding neural elereinforce the thoughts of ments.” These statements Fine and Brucker in connection with their finding of macular microinfarction: “Such vascular changes, of course, can produce areas of retinal ischemia, to which the Miiller cells in this region of the foveomacula would be vulnerable. This vulnerability may be related to the morphologic observation that foveomacular Miller cells occupy a greater volume and have a more uniform lucency of cytoplasm than they do elsewhere in the retina.“” In contrast to the paucity of investigations dealing specifically with retinal edema, experimental brain edema has been extensively studied, probably because of the life threatening circumstances associated clinically with brain edema. A recent volume of Advances in Neurology (1980) entitled Brain Edema provides considerably background for investigating and understanding retinal edema. What the neuroscientists term “vasogenic brain edema” is of particular interest to us, and in the volume noted above, Klatzo et al” authored a most pertinent article on the resolution of vasogenic brain edema. Based on a series of experiments, they proposed the following hypothesis concerning development and resolution of brain edema. In vasogenic brain ede-
ANIMAL
MODELS OF MACULAR
523
EDEMA
ma, injured vessels allow edema fluid and serum proteins to enter the extracellular space. Astrocytes are swollen by edema fluid and proteins; the subsequent intracellular lysomal digestion of the proteins then provides the basic mechanism for removing proteins from the extracellular space. This protein removal allows a return to normal of colloidal osmotic forces which in turn allows water to move out of the extracellular space across the blood-brain barrier.” Thus, edema resolution is hypothesized to be dependent upon removal of osmotically active proteins by intraglial uptake of the extravasated serum proteins. Since the principal glial cell of the retina is the Miiller cell, we may also hypothesize that prevention and resolution of retinal edema is dependent upon Miiller cell function (with, of course, appropriate return to normal of the leaky blood-retinal barriers). Thus, if blood-retinal barrier function is reestablished and Miiller cell function remains operative, macular edema may not occur. Tso’s ocular hypotony model may represent this in that (1) leakage without edema occurred for a short period of time, (2) HRP was observed in Miiller cells, and (3) shortly afterward the blood-retinal barriers no longer leaked proteins. In essence, this is an edema prevention model. Retinal edema was also not observed in the aphakit edema model even though leakage of proteins was demonstrated during the 30-day postoperative period. But again the edema resolution mechanisms (the Miiller cell?) were apparently capable of preventing the development of macular edema. In the photic maculopathy model the acute edematous process was resolved, but over a period of almost four years, a persistent leak ofserum proteins may have begun to tax the Miiller cell such that mild edema of the outer plexiform (Henle’s) layer became evident. We do not know if that edema was extracellular, intracellular or if it involved the Miiller cell. But, as we can now see, these questions are important. On the other hand, after a period of three to four years, the argon laser maculopathy model showed vacuolation (edema?) of the extracellular space in Henle’s layer. There was also evidence of proteinaceous material in the enlarged (cystoid?) extracellular spaces. This model could therefore represent a mild but chronic leakage of proteins across the blood-retinal barriers and an impaired ability of the Miiller cell to take in the extravasated proteins. It is possible, for example, that the laser had destroyed a significant portion of the macular component of Miiller cells. Talc retinopathy with macular edema, as noted earlier, is both interesting and informative. In that
situation we have a multiple and prolonged insult to the blood-retinal barriers, producing sufficient impact upon the neural retina to cause a number of retinal layers to develop cystoid edema. It would seem that the extracellular cystic spaces are in contrast to the intracellular (Miiller) cystoid edema described in the human cases, but even so, this model presents an important opportunity to study development and nonresolution of macular edema. We can now see that Tso’s statement in his Cambridge Symposium paper was indeed prophetic, that indeed pathogenetic mechanisms other than simple disruption of the blood-retinal barriers are necessary for the development of cystoid macular edema.” I sue;qest that Miiller cell dysfunction may IL he one of those mechanisms. There certainly are other mechanisms responsible for movement of fluid and other substances through the retinal tissue (e.g., the Na+ K’ ATPase pump) and undoubtedly it is necessary to investigate these biochemical mechanisms as well. At the same Cambridge Symposium, I reported on work in our laboratory and by other investigators that indicated the importance of the retinal environment upon the morphologic characteristics of developing and/or existing retinal vessels.’ It would appear, based upon our present concepts that we may need to manipulate the retinal environment as well as disrupt the blood-retinal barriers in order to develop models of chronic cystoid macular edema. That is, we need to create an environment that is not conducive to the resolution of the edematous process. There are various ways that this could be accomplished. For example, permeability abnormalities in the blood-retinal barrier have been demonstrated in monkeys with experimental hypertension” or diabetes.” If there were a macular insult superimposed on these abnormal retinas, would a cystoid macular edema result? .4nother approach would he to use selected cytotoxic agents to impair edema resolution mechanisms. For example, sodium iodate damages the integrity of the retinal pigment epithelial cells, oubain interferes with Na+ K+ XTPase pump mechanisms, alpha amino-adipic acid can damage the Miller cell, etc. Can we, through the use of agents such as these, selectively produce various types of retinal edema? I believe that we must if we are to develop suitable animal models which will enable us to understand the mechanisms responsible for the development, resolution and nonresolution of macular edema.
References
524
Surv Ophthalmol
28(Suppl)
May
1984
mol 6% Sci 19(ARVO .Suppl):4, 1980 3. Fine BS, Brucker A.J: Macular edema and cystoid macular edema. AmJ Ophthalmol 92:466-481, 1981 4. Henkind P, Bellhorn RW, Schall B: Retinal edema: postulated mechanism(s), in Cunha-Vaz.JG (ed): The Blood-Retinal Barriers. New York, Plenum Publishing, 1980, pp 251-268 5. Irvine AR: Cystoid maculopathy. Sure Ophthalmol21: I-1 7, 1976 6. ,Jampol LM. Setogawa T, Rednam KRV. Tso MOM: Talc retinopathy in primates. A model of ischemic retinopathy: I. Clinical studies. Arch Ophthalmol 99: 1273-1280, 1981 7. Kaga N, Tso MOM. Jampol LM, et al: Talc retinopathy in primates: A model of ischemic retinopathy. II. A histopathologic study. Arch Ophthalmol 1011:1644-1648, 1982 8. Kaga N. Tso .MOM,Jampol LM: Talc retinopathy in primates: A model of ischemic retinopathy. 111. An electron microscopic study. Arch Ophthalmol 100: 1649-1657, 1982 9. Klatzo I, Chui E, Fu,jiwara K, Spatz M: Resolution ofvasogenic brain edema, in Cervos-Xavarro .J. Fcrszt R (cds): .4dza:a,ws in .Ye’eurologv /[‘al 28) Brain Edema. NW York. Raven Press, 1980, pp 359-373 IO. Kohner EM, Dollcry CT. Shakib Xl. ct al: Experimental retinal branch vein occlusion. rim J Ophthaltncl 6Y:778-825, 1970 11. Tso MOM: Pathological study of cystoid macular ordema. Tram Ophthalmol Sac (‘I( 100:408-413, 1980 12. Tso X1011: Pathology of the blood-retinal barrier, in CunhaVaz,JG (rd): T/w Blood-Rebnal Barriers. New York, Plenum Puhlishing, 1980, pp 235-250
BELLHORN 13. ‘l’s0 MOM: Pathology of cystoid macular edema. Ophthalmology R9:902-9 14, I982 14. Tso MOM: Photic maculopathy in rhesus monkey. A light and electron microscopic study. InvestOphthalmol 22: 17-34, 1973 15. Tso MOM: Cunha-Vaz J, Blair N, et al: Disruption of bloodretinal harrier in diabetic primates. Invest Ophthalmol Vis Sci -32/8RVO Suppl): 110, 1982 16. Tso MOM, Fine BS: Repair and late degeneration of the primate foveola after injury by argon laser. InueslOplrlhalmof Vis Sri l&447-461, 1979 17. ‘l‘s0 MOM, Fine BS, Z’lmmerman LE: Photic maculopathy produced by the indirect ophthalmoscope. I. Clinical and histopathologic study. Am J Ophthalmoi 73:68&699, 1972 18. Tso MOM. Shih CY: Disruption of blood-retinal harrier in ocular hypotony: Preliminary report. Exp $ve Res 23:20%216, 1976 19. ‘I’s0 MOM, Shih CY: Experimental macular edema after lens extraction. lnmt Ophthalmof Vis Sri 16:381-392, 1977
Supported by NE1 Grant EY 02038 and hy Research to Prevent Blindness, Inc. Reprint requests should be addressed to Roy Bellhorn, D.V.M., hlhrrt Einstein College of Medicine, I I1 East 210th Street, Bronx, NY 10467.