Graefe's Arch Clin Exp Ophthahnol (1991) 229:568-573 Graefe's Archive for Clinical and Experimental

Ophthalmology © Springer-Verlag 1991

Experimental traction retinal detachment in the cat* Charles A. Wilson, Joseph A. Khawly, Diane L. Hatchell, and Robert Machemer

Durham Veterans Administration Medical Center and Duke University Eye Center, Box 3802, Durham, NC 27710, USA

Received November 9, 1990 / Accepted January 21, 1991

Abstract. We developed a reproducible model of traction retinal detachment (TRD) in the cat eye by creating a serous retinal detachment and then injecting 2.5 x 105 kitten dermal fibroblasts into the vitreous cavity at the site of a retinal wound. Serous detachments were pro- duced by exposing an area of retina to focused light after intravenous injection of rose bengal (a photosensi- tizing dye). TRD developed rapidly within the first 2 weeks after fibroblast injection, accompanied by the formation of vitreoretinal strands and, to a lesser degree, epiretinal and/or subretinal proliferation. Histopatho- logy demonstrated fibroblasts within the vitreous or along the posterior hyaloid face. Focal deposits of fibro- blasts were occasionally found on the inner surface of the retina and/or in the subretinal space. Fibroblast pro- liferation was confirmed by uptake of radiolabeled thy- midine. Deposition of collagen was noted at as early as 3 days after fibroblast injection. Neovascularization was not observed. Control eyes that did not receive fibroblasts showed resolution of serous detachment without retinal traction. In all eyes, retinal degeneration and thinning were seen in the area of previous photody- namic treatment. In this model of TRD, anteroposterior traction (due to vitreous strands) predominates, as is observed in experimental posterior penetrating ocular injury induced by intravitreal blood injection, which also results in vitreous strand formation. Our model, how- ever, enables clinical assessment of TRD in the cat with- out the media opacification produced by vitreous blood.

Introduction

Tractional detachment (TRD) is the second most com- mon form of retinal detachment [12]. TRD occurs when forces either perpendicular or tangential to the retinal

* Supported by VA medical research funds, NIH research grant EY02903, core grant EY05722, the Helena Rubinstein Foundation, New York, and Research to Prevent Blindness, Inc., New York. Dr. D.L Hatchell is a Research to Prevent Blindness, Inc., Senior Scientific Investigator. The authors have no commercial or proprie- tary interest in the chemicals, drugs, or devices used in this study

Offprint requests to: D.L. Hatchell

surface become sufficient to overcome adherence of the retina to the pigment epithelium. This can occur in a variety of conditions, including proliferative diabetic re- tinopathy, proliferative vitreoretinopathy, retinopathy of prematurity, and posterior penetrating eye injury. Each of these conditions is an important cause of visual disability or blindness.

Several animal models have been developed to ad- vance the study of TRD. Machemer and Norton [16], for example, described a process resembling proliferative vitreoretinopathy (then called massive periretinal prolif- eration) in monkeys with chronic rhegmatogenous reti- nal detachments. Subsequent studies revealed the cellu- lar nature of epiretinal proliferations in these eyes [13- 15]. A simpler, albeit more artifical, method of produc- ing TRD was developed by Algvere and Kock [1], who injected large numbers of autologous fibroblasts into the vitreous cavity of rabbit eyes and produced spontaneous traction detachment in a high percentage of animals. Refinements of their method have included the use of fewer numbers of homologous fibroblasts and simula- tion of posterior vitreous detachment [3, 8, 17, 18]. TRD has also been induced following posterior penetrating eye injury in rabbits [4, 5, 19, 20], pigs [9, 10], and monkeys [6]. In many of these studies, however, prolifer- ation either was mild or did not occur in the absence of vitreous blood [4-6, 20]. After traumatically induced vitreous hemorrhage or intravitreal injection of autolo- gous blood, proliferation was enhanced, but observation of the fundus was impaired.

In this report we describe a reproducible method of producing TRD in the cat eye. The detachments resem- ble the vitreoretinal fibrosis associated with posterior penetrating ocular trauma, since they predominantly in- volve anteroposterior traction. Severe vitreous hemor- rhage is avoided and the ocular media remains clear, enabling close observation of vitreoretinal proliferation.

Materials and methods

Animals

A total of 16 cats of either sex, weighing between 3 and 6 kg, were used in this study. The animals were maintained and treated

Table 1. Results of injections of fibroblasts into the vitreous cavity of cat eyes as compared with controls

Cat Clinical Length of Autoradio- Histo- outcome follow-up graphy pathology

(days)

Fibroblast- 1 SRD 1 Yes Yes injected eyes 2 SRD 3 Yes Yes

3 SRD 6 Yes Yes 4 TRD 9 No No 5 TRD 14 No Yes 6 TRD 21 No Yes 7 TRD 21 Yes Yes 8 TRD 28 No No 9 TRD 28 No Yes

10 TRD 28 No Yes i 1 TRD 42 No Yes 12 TRD 56 No No

PBS-injected 1 RRD 18 No Yes eyes" 2 RRD 18 No No

3 RRD 28 No No 4 RRD 28 No Yes

a Received no fibroblasts SRD, Serous retinal detachment; TRD, retinal detachment with traction; RRD, resolved retinal detachment without traction; PBS, phosphate-buffered saline

in accordance with the Association for Research in Vision and Ophthalmology (ARVO) resolution on the use of animals in re- search. Institutional guidelines were followed throughout the study. During all procedures, the cats were anesthetized by intramuscular injection of ketamine hydrochloride (20 mg/kg) and acepromazine maleate (0.5 mg/kg). The pupils were dilated using topical pheny- lephrine hydrochloride (5%) and tropicamide (0.25%). Indirect ophthalmoscopy and color fundus photography were performed before and after all procedures as well as on each day of follow-up. Follow-up examinations were performed weekly and terminated at the times specifed in Table 1. Cats were killed by intravenous injection of pentobarbital sodium (100 mg/kg) for histologic study on days 1, 3, 6, 14, 21, 28, and 42. Autoradiography was performed on eyes obtained on days 1, 3, 6, and 21, which had received an intravitreal injection of [3H]-thymidine (0.05 mCi; specific activity, 740 GBq/mmol; volume injected, 50 gl) 24 h prior to euthanasia.

Production of retinal detachment

Serous retinal detachment was produced in one eye of each animal using a photodynamic technique based on previous studies [21]. Briefly, rose bengal was given intravenously at a dose of 10 mg/kg, after which it was photochemically activated in the eye using a xenon photocoagulator (Clinitex, Varian Associates Inc., Danvers, Mass.) filtered to a central wavelength of 550_+10 nm. The irra- diance (measured at the surface of the cornea) was 30 mW/cm 2 and the time of exposure was 3 min. A circular area of superonasal retina (over the tapetum) measuring approximately 6 mm in diame- ter was exposed through a fundus contact lens. Serous retinal de- tachments developed rapidly in all eyes, enlarging over the next 2 days to involve most of the nasal retina.

Fibroblast cultures

Fibroblasts were obtained from a single line of cells cultured from the flank dermis of a single newborn kitten. Culture techniques were identical to those reported by Chandler et al. [3] for adult rabbit dermal fibroblasts. Prior to use, the fibroblasts were sus-

569

pended in phosphate-buffered saline at a concentration of 250000 cells/0.1 ml; they were injected into eyes within 2 h of their removal from the culture medium.

Cell injection

On the 2nd day after photodynamic treatment, a fornix-based con- junctival peritomy and pars plana sclerotomy were performed in the superior-temporal quadrant, through which a vitreous cutter was passed. Using an operating microscope and irrigating contact lens, the cutting instrument was positioned over the detached nasal retina. A retinal hole measuring 1-3 mm in diameter was then created in an area devoid of large retinal vessels. A small amount of retinal bleeding was occasionally noted after this procedure, but severe vitreous hemorrhage was not encountered. The cutter was then used to remove a small amount of vitreous around the retinal hole.

In 12 eyes, homologous kitten dermal fibroblasts (250000 cells in 0.1 ml phosphate-buffered saline) were injected at the site of the retinal hole using a blunt 23-gauge needle. During the injection, cells were observed to enter both the subretinal space and the vitreous cavity. Another 4 eyes (controls) were injected with an identical volume of phosphate-buffered saline containing no fibro- blasts. After injection, the sclerostomy was closed using an 8-0 suture and erythromycin ointment (5%) was applied.

Histology

All experimental eyes were fixed for at least 3 days in 10% neutral buffered formalin, then opened and embedded in paraffin. Sections were obtained in a stepwise fashion throughout each eye and were stained using Masson's trichrome stain. Adjacent sections in se- lected eyes were obtained for autoradiography. These sections were deparaffinized, coated with Kodak NTB-2 emulsion, and exposed at room temperature for 8 days. Sections were then developed in D-19 developer (diluted 1:1, v/v) for 4 rain, fixed, and stained with Mayer's hematoxylin. One control eye was fixed over 1 week in a 1 : 1 (v/v) solution of 4% paraformaldehyde and 5% glutaralde- hyde. The posterior segment was then dissected, fixed for 90 min in 2% buffered osmium tetroxide, and embedded in a low-viscosity epoxy resin. Sections were stained with methylene blue and basic fuchsin.

Results

Clinical findings

In all, 9 of 12 exper imental eyes were observed for at least 1 week after f ibroblas t injection. These eyes devel- oped clinical signs of T R D , consis t ing pr imar i ly of vi- t reoret inal s t rands and, to a lesser extent, subret inal and epiretinal prol i ferat ions (Fig. I). Vi t reoret inal s tands were usual ly a t tached to the edges of ret inal wounds and extended anter ior ly toward the lens or pars plana. In the major i ty of eyes, the s trands extended slightly inferiorly ra ther than superiorly toward the sclerotomy site. Occasionally, vi t reous s t rands were found over other areas of detached retina, in which case they formed a lattice-like veil wi thin the inferior vitreous and did no t appear to a t tach directly to the retina.

Vitreous haze occurred postoperat ively in mos t ani- mals bu t did no t obscure the view of the re t ina; in all cases, this haze had cleared by the end of the 1st week. Retinal t rac t ion increased rapidly dur ing the first 1-

570

571

3

Fig. 3. Thymidine uptake at 1 day after fibroblast injection. A mass of fibro- blasts in the subretinal space contains proliferating cells distributed primarily in the periphery. Mayer's hematoxylin, original magnification x 400

Fig. 6. Histologic appearance of an eye not subjected to fibroblast injection (28 days following injection of PBS). Retinal reattachment is evident along with marked thinning and gliosis of the retina in the area of photodynamic treatment. A sharp demarcation is visi- ble between the treated and untreated retina. Methylene blue-basic fuchsin, original magnification x 66

Fig. 1 a, b. Progressive development of traction retinal detachment and vitreoretinal fibroplasia, a At day 2 after fibroblast injection, proliferations are seen in small clumps in the vitreous (arrows) in the vicinity of the newly created retinal hole (arrowhead). A small amount of traction has occurred, as evidenced by the irregu- lar folding and tenting of the retina, b At day 42, fibroblasts have formed strands (arrows) that attach to the edge of the retinal wound (arrowhead). The wound and retina show evidence of further trac- tion

Fig, 2. Strand-shaped proliferation overlying the area of the optic disc at 3 days after fibroblast injection. The strand appears highly cellular, showing a small area of collagen deposition which stains bluish-green (arrow). Masson's trichrome, original magnification × 66

Fig. 4a, b. Day 14 after fibroblast injection, a A mass of fibroblas- tic proliferation is noted at the edge of the retinal wound (arrow- head), and some cells extend anteriorly along the detached posterior hyaloid face (arrows). Masson's trichrome, original magnifica- tion x 3.3. b Higher magnification shows extensive staining of colla- gen and attachment to the edge of the gliotic retina (arrow). Mas- son's trichrome, original magnification × 13.2

Fig. 5. Day 21 after fibroblast injection. An epiretinal mass of fi- broplasia is seen extending (arrow) onto the detached posterior hyaloid face. Cellularity has decreased, and collagen occupies the majority of the mass. Masson's trichrome, original magnification × 13.2

572

2 weeks and slowly thereafter. New retinal tears fre- quently developed in the periphery of eyes that were followed over 3 weeks. In one eye, a mild vitreous hem- orrhage was observed on day 21 that had not been pres- ent initially. In two eyes, peripheral retinal reattachment had developed by day 14, but severe retinal traction re- mained in the peripapillary area. Retinal folding had obscured any view of the retinal wound in both eyes. Four control eyes injected with phosphate-buffered sa- line containing no fibroblasts showed complete retinal reattachment without folds or traction during a follow- up period of either 18 (two eyes) or 28 days (two eyes).

Histologic findings

In all 12 experimental eyes, fibroblasts were noted in one or more of the following locations: beneath the reti- na, in the vitreous (clumps or strands), along the posteri- or myoloid face, or on the inner retinal surface (Figs. 2- 5). On days 1 and 3 after cell injection, small collections of fibroblasts were noted in the vitreous cavity or subret- inal space (Fig. 2). On day 3, these collections showed a small amount of staining for collagen. At as early as 1 day after fibroblast injection, proliferative activity was indicated by thymidine uptake in both fibroblasts (Fig. 3) and the inner retina. Thymidine uptake by fibro- blasts was also noted on days 3 and 6, and was usually seen in cells at the periphery of the fibroblast deposits. Associated collagen staining increased over time.

At 14 days after cell injection, a further increase in the size of fibroblast collections and associated collagen staining was noted (Fig. 4). Fibroblasts were occasional- ly found along the separated posterior hyaloid face, in which case they followed lines of anteroposterior retinal traction. Some areas of retina showed degenerative changes consistent with ischemia due to photodynamic treatment and subsequent vascular thrombosis. On day 21, fibroblast deposits appeared to be less cellular and richer in collagenous ground substance (Fig. 5). Those that had settled on the inner surface of the retina consisted of globular deposits attached at one side to the retina. No extension was seen along the inner retinal surface. Subretinal deposits, however, tended to spread laterally along the pigment epithelial layer or Bruch's membrane. Pigment-containing cells were admixed with fibroblasts in the vitreous cavity. Thymidine incorpora- tion was absent. By 28 days after fibroblasts injection, no further histologic changes had occurred.

Eyes that did not receive fibroblasts showed complete retinal reattachment at either 18 or 28 days (Fig. 6). No fibroblast proliferation or retinal traction was found in these eyes. Again, the retina in the area of photodynamic treatment was thinned and atrophic and a sharp demar- cation was noted at its edge, outside which the retina appeared normal.

Discussion

TRD was consistently produced in cat eyes without the need for intravitreal blood injection. The cat was chosen

for these experiments because its retinal architecture is closer to that found in primates, the eye is larger, and the inflammatory response to irritative stimuli is consid- erably weaker than that in the rabbit [2]. Typical clinical and histopathologic signs of intraocular proliferation de- veloped, consisting of vitreoretinal strands, proliferation along the posterior vitreous face, and a few subretinal fibroblast deposits. However, preretinal proliferation in the form of epiretinal membranes was not observed. The proliferating cells exhibited thymidine uptake during the 1st week, and a progressive increase in collagen staining was observed over 2-3 weeks. Although proliferative ac- tivity was not quantified in this study, its timing ap- peared to be similar to that previously reported in the fibroblast-injected rabbit eye, in which a reactive prolif- eration in the retina was also noted [11].

A preexisting retinal detachment and the presence of fibroblasts in the vitreous cavity seemed to be necessary for the development of TRD. The contribution of the retinal hole in this process was less clear, but it appeared to enable or promote vitreoretinal proliferation. In pre- liminary studies, serous retinal detachment alone or in combination with fibroblast injection (without a retinal hole) failed to produce TRD. In the present model, the edge of the retinal wound was frequently the site of vi- treoretinal proliferation. Proliferation along the inner retinal surface did not occur and may have been pre- vented by a firmly attached cortical vitreous. This as- sumption is supported by additional preliminary experi- ments in which we injected 5 x 105 kitten dermal fibro- blasts into the intact vitreous of eyes with attached reti- nas. The cells failed to settle on the retina and either disappeared or migrated anteriorly. No TRD or epiretin- al membrane was seen (unpublished results). Such find- ings suggest that creation of a retinal hole either may enable fibroblasts gain access to the retina by focally disrupting the cortical vitreous or may have additional importance related specifically to retinal trauma and its effect on cell adherence or proliferation.

Erickson et al. [7] did not report the development of tractional proliferation in cat eyes with rhegmatogen- ous retinal detachments lasting as long as 14 months, despite the growth of subretinal membranes derived from Miiller cell processes. To produce long-lasting de- tachments, these authors performed both a vitrectomy and a lensectomy before creating a retinal hole. In our study, "rhegmatogenous" detachment was induced by placing a hole in an area of preexisting serous retinal detachment; in this way, we avoided the need for more extensive surgical procedures. Fortuitously, the retinal reattached spontaneously in all control eyes, in contrast to experimental (fibroblast-injected) eyes, in which reti- nal elevation persisted or increased throughout the study.

In some models of TRD, epiretinal membrane forma- tion has been observed that has been compared to prolif- erative vitreoretinopathy in humans [3, 5, 6, 13-15]. In our study, however, epiretinal membranes did not devel- op. Instead, proliferation and traction occurred in an anteroposterior direction because of the formation of transvitreal strands that attached at the edge of the reti-

573

nal wound. This appearance more closely resembles that of perforating ocular t rauma with transvitreal fibrosis [19].

We reviewed the literature on perforating ocular trau- ma and found potential advantages of the present model over previous animal models of this condition. Topping and co-workers [19] induced double-perforating injury in rabbit eyes and observed strand-shaped fibroblastic proliferations following the vitreous injury track; they found little or no retinal detachment despite this prolifer- ative response. Gregor and Ryan [9, 10] performed com- bined posterior contusion and penetrating injuries in pig eyes and found similar transvitreal traction bands; again, the resulting retinal detachments were generally small. In addition, proliferation depended on the injec- tion of relatively large quantities of blood into the vitre- ous cavity, which obscured clinical observation of these eyes. Blood injection was also found to be an essential factor in the development of traction in rabbit eyes with experimental penetrating injury [4, 5, 20]. In monkeys, comparable methods did not produce anteroposterior vitreous traction because of the development of posterior vitreous detachment [6]. However, in some animals the peripheral retina was dragged toward the pars plana, resulting in retinal detachment, a condition that required over 6 weeks to develop. More recently, Yeo et al. [22] created T R D in rabbits using a standard posterior pene- trating injury and intravitreal injection of platelet-de- rived growth factor and fibronectin. This procedure avoided the previous problem of decreased media clear- ity due to vitreous blood. However, because of the strik- ing dissimilarity of the rabbit and human retina, this model is limited in its translation to human TRD.

The present model was developed in the cat eye, which has a holangiotic (fully vascularized) retina like that of humans. Retinal traction was produced rapidly and reproducibly as opposed to the slow evolution of T R D in the aforementioned pr imate model [6]. This ra- pid clinical course may be explained as follows: first, by the injection of fibroblasts directly into the vitreous, the process of natural invasion of the vitreous injury track by fibroblastic cells is bypassed; and second, less traction (or proliferation) should be required to hold the retina in a detached position than would be necessary for its pr imary detachment f rom the retinal pigment epi- thelium.

In summary, we developed a reproducible model of T R D in the cat eye, which is produced by transvitreal fibroblast injection at the site of a wound in a previously detached retina. The model results in a clinical picture of T R D resembling that seen in penetrating ocular trau- ma but avoids the need for injection of blood into the vitreous and hence, media opacification.

References

1. Algvere P, Kock E (I 976) Experimental fibroplasia in the rabbit vitreous: retinal detachment induced by autologous fibroblasts. Graefe's Arch Clin Exp Ophthalmol 199 : 215-222

2. Bito LZ (1984) Species differences in the responses of the eye to irritation and trauma: a hypothesis of divergence in ocular defense mechanisms, and the choice of experimental animals for eye research. Exp Eye Res 39: 807-829

3. Chandler DB, Quansah FA, Hida T, Machemer R (1986) A refined experimental model for proliferative vitreoretinopathy. Graefe's Arch Clin Exp Ophthalmol 224:86-91

4. Cleary PE, Ryan SJ (1979) Experimental posterior penetrating eye injury in the rabbit: I. Method of reproduction and natural history. Br J Ophthalmol 63:306-311

5. Cleary PE, Ryan SJ (1979) Experimental posterior penetrating eye injury in the rabbit: II. Histology of wound, vitreous, and retina. Br J Ophthalmol 63:312-321

6. Cleary PE, Ryan SJ (1979) Histology of wound, vitreous and retina in experimental posterior penetrating eye injury in the rhesus monkey. Am J Ophthalmol 88:221-231

7. Erickson PA, Fisher SK, Anderson DH, Stern WH, Borgula GA (1983) Retinal detachment in the cat: the outer nuclear layer and outer plexiform layers. Invest Opthalmol Vis Sci 24: 927-942

8. Fastenberg DM, Diddie KR, Dorey K, Ryan SJ (1982) The roie of cellular proliferation in an experimental model of massive periretinal proliferation. Am J Ophthalmol 93:565- 572

9. Gregor Z, Ryan SJ (1982) Combined posterior contusion and penetrating injury in the pig eye: I. A natural history study. Br J Ophthalmol 66:793-798

10. Gregor Z, Ryan SJ (1982) Combined posterior contusion and penetrating injury in the pig eye: II. Histologic features. Br J Ophthalmol 66 : 799-804

11. Hatchell DL, McAdoo T, Sheta S, King RT III, Bartolome JV (1988) Quantification of cellular proliferation in experimen- tal proliferative vitreoretinopathy. Arch Ophthalmol 106:669- 672

12. Landers MB III, Hjelmeland LM (1989) Types of pathogenetic mechanisms of retinal detachment. In: Ryan SJ (ed) Retina vol 3. Mosby, St. Louis, pp 105-109

13. Laqua H, Machemer R (1975) GliaI cell proliferation in retinal detachment (massive periretinal proliferation). Am J Ophthal- mol 80:602-618

14. Laqua H, Machemer R (1975) Clinical-pathological correlation in massive periretinal proliferation. Am J Ophthalmol 80:913- 929

15. Machemer R (1968) Experimental retinal detachment in the owl monkey: II. Histology of retina and pigment epithelium. Am J Ophthalmol 66 : 396-410

16. Machemer R, Norton EWD (1968) Experimental retinal de- tachment in the owl monkey: I. Methods of production and clinical picture. Am J Ophthalmol 66:388-396

17. Ophir A, Blumenkranz MS, Claflin AJ (1982) Experimental intraocular proliferation and neovascularization. Am J Oph- thalmol 94:450-457

18. Sugita G, Tano Y, Machemer R, Abrams G, Claflin A, Fioren- tino G (1980) Intravitreal autotransplantation of fibroblasts. Am J Ophthalmol 89:121-130 Topping TM, Abrams GW, Machemer R (1979) Experimental double-perforating injury of the posterior segment in rabbits eye. The natural history of intraocular proliferation. Arch Op- thalmol 97:735 742 Vergara O, Ogden T, Ryan S (1989) Posterior penetrating injury in the rabbit eye: effect of blood and ferrous ions. Exp Eye Res 49:1115 1126 Wilson CA, Royster AJ, Tiedeman JS, Hatchell DL (1991) Exu- dative retinal detachment following photodynamic injury. Arch Ophthalmol 109:125-134 Yeo JH, Sadeghi J, Campochiaro PA, Green WR, Glaser BM (1986) Intravitreous fibronectin and platelet-derived growth factor. New model for traction retinal detachment. Arch Oph- thalmol 104:417-421

19.

20.

21.

22.