- Email: [email protected]
Experimental Transvitreal Cyanoacrylate Retinopexy Through Silicone Oil
Experimental Transvitreal Cyanoacrylate Retinopexy Through Silicone Oil Sherif M. Sheta, M.D., Tetsuo Hida, M.D., and Brooks W. McCuen II, M.D.
We evaluated the use of transvitreal cyanoacrylate retinopexy in the treatment of experimental rhegmatogenous retinal detachment during vitreous surgery in rabbit eyes filled with silicone oil. The view to the fundus was superior to that obtained in our previous model of cyanoacrylate retinopexy in the airfilled eye. Glue delivery was consequently both easier and more precise through silicone oil relative to air. The chorioretinal adhesions produced with cyanoacrylate tissue adhesive were compared with those produced by transscleral retinal cryopexy and were found to be more rapid in onset as well as stronger. An exaggerated tissue response adjacent to the cyanoacrylate site suggested a potential toxic chemical or thermal reaction, or both, to the tissue adhesive, but there was no evidence of any distant ocular effects. WE HAVE PREVIOUSLY described a model of rhegmatogenous retinal detachment during vitreous surgery in the rabbit with intraoperative repair by fluid-air exchange and the transvitreal application of n-butyl-cyanoacrylate tissue adhesive through the air-filled eye to the retinal break. 1 The chorioretinal adhesions achieved were immediate in onset, stronger than those produced with transscleral retinal cryopexy, and long-lasting in effect. In this study we have modified our experimental model so that cyanoacrylate retinopexy is performed through silicone oil-filled eyes rather than in eyes filled with air.
Accepted for publication Sept. 22, 1986. From the Department of Ophthalmology, Duke University Eye Center, Durham, North Carolina. This study was supported by the Adler Foundation, the National Society to Prevent Blindness, and the Peace Fellowship Program. Reprint requests to Brooks W. McCuen II, M.D., Box 3802, Duke University Eye Center, Durham, NC 27710.
©AMERICAN JOURNAL OF OPHTHALMOLOGY
102:717-722,
Material and Methods Twenty-eight pigmented rabbits weighing from 2.5 to 3.5 kg were used in this study. Two weeks before vitrectomy the planned sclerotomy sites 2 mm posterior to the corneoscleral limbus in the superotemporal, superonasal, and inferotemporal quadrants were marked with 6-0 silk sutures and treated with prophylactic transscleral retinal cryopexy. Three days before vitreous surgery, by indirect ophthalmoscopy, 0.2 cc of pure perfluoropropane gas was injected into the mid-vitreous cavity with a 30-gauge needle through the inferior cryopexy site. The animal was placed upright to prevent the gas from escaping through the injection site. After the intraocular pressure fell below 20 mm Hg, the eyes were reinjected with an additional 0.2 cc of perfluoropropane. One hour before vitreous surgery pupillary dilatation was achieved with topical 0.5% phenylephrine hydrochloride, 0.25% tropicamide, and 1% atropine sulfate. The animals were anesthetized with intramuscular injection of 1.0 ml of ketamine hydrochloride (100 mg/ml) and 0.1 ml of xylazine (100 mg/ml). After retrobulbar injection of 1.0 ml of 1% lidocaine hydrochloride, a conjunctival peritomy was performed for 360 degrees at the corneoscleral limbus. The previously marked sclerotomy sites were identified and a preplaced 7-0 Dexon mattress suture was passed around the inferotemporal site. A 1.4-mm microvitreoretinal blade was used to enter the vitreous cavity at the inferotemporal site and a 2.5-mm infusion cannula delivering Ringer's lactate solution with 5% dextrose was secured in position with the preplaced mattress suture. Similar sclerotomies were performed in the superotemporal and superonasal quadrants. By initiating infusion, a total gas-fluid exchange with refilling of the eye with Ringer's lactate was achieved. A bent 19-9auge butterfly infu-
DECEMBER,
1986
717
718
December, 1986
AMERICAN JOURNAL OF OPHTHALMOLOGY
sion needle was passed through the superonasal sclerotomy while the Shock phacofragmentor needle was inserted through the superotemporal sclerotomy to fragment and aspirate the lens nucleus. The lens cortex and the anterior and posterior lens capsules were removed with the I-mm vitreous cutter. In five cases we left the anterior capsule intact to prevent silicone oil from filling the anterior chamber during later fluid-silicone exchange. After replacing the 19-9auge butterfly needle with the fiberoptic tissue manipulator.! an anterior and posterior vitrectomy was performed with the vitreous cutter. After removing as much cortical vitreous as possible, the vitreous cutter was directed toward the retina, and two retinal holes approximately 500 fL in diameter were created 1 disk diameter above and below the optic disk. An infusion stream of Ringer's lactate from the fiberoptic probe was directed over the retinotomy sites to balloon up the subretinal space and create localized posterior retinal detachments. A fluid-silicone exchange procedure with internal drainage of the subretinal fluid through the iatrogenic retinotomies using a blunt 19-9auge needle then effected retinal reattachment and an eye filled with silicone oil. At this time the retinotomies were randomized so that in each eye one retinotomy was treated with transvitreal cyanoacrylate retinopexy while the other retinotomy was treated with transscleral cryopexy. Cyanoacrylate retinopexy was performed with a 50:50 mixture of n-butyl-2-cyanoacrylate and icphendylate using the specially designed glue applicator previously described." The glue injector was held close to the retinal hole, and a tiny drop of glue was produced at its tip by gentle finger pressure on the actuator. The injector tip with the small bubble of tissue adhesive was brought into contact with the retinotomy site and was then quickly retracted to prevent the injector tip from adhering to the retina during the polymerization of the adhesive. Transscleral retinal cryopexy was performed under microscopic control after closure of the two superior sclerotomies with 7-0 Dexon sutures. A bright white choroidal and retinal freeze was produced to surround the retinotomy while making sure that the cryopexy reaction occurred only at the retinotomy site and that it approximated the size of the cyanoacrylate application (approximately 1,000 u.). The infusion cannula was removed and its scleroto-
my closed with a preplaced mattress suture. Ten milligrams of gentamicin was injected into the subconjunctival space, and 1% atropine and dexamethasone and polymixin B sulfate ointments were applied topically to the globe. Seven eyes were enucleated immediately after surgery and the adhesive forces were measured using retinal peeling at the retinotomy site.! In seven eyes ophthalmoscopic and slit-lamp examinations were performed on days 2, 4, and 7, after which the eyes were enucleated and adhesive force measurements were performed. The 12 eyes in the third group were examined at days 2, 4, 7, and weekly thereafter for one month, at which time the eyes were enucleated and adhesive force measurements were performed. An additional eye was enucleated 24 hours after surgery and fixed in 10% buffered formalin. Selected specimens were stained with hematoxylin and eosin and oil red 0 stains for histologic evaluation. In one other eye we used pure cyanoacrylate without iophendylate to seal the retinal hole. Statistical analysis of the chorioretinal adhesive forces was done after logarithmic transformation of the original measure of force since the coefficient of variation remained constant over a wide range of means. The log transformed data were analyzed using a repeated measures two-way analysis of variance design. The two main factors were the method of retinopexy and the time period of the peeling."
Results We encountered several technical problems associated with cyanoacrylate retinopexy in the silicone oil-filled eyes. During cyanoacrylate delivery the tissue adhesive polymerizes more slowly under silicone oil than under air. Although slower polymerization could lead to undesired spreading of glue onto the adjacent retina, this was prevented by using only small amounts of tissue adhesive. Additionally, the glue has a specific gravity less than that of silicone oil and the cyanoacrylate drop tends to float upward from the retinal surface. This tendency was counteracted by lightly pushing the bubble of glue downward against the floor of the retinal break until polymerization began to occur. Also, visualization of the transparent glue drop was difficult in the silicone-filled eye because of the similarity of their respective
Vol. 102, No.6
Transvitreal Cyanoacrylate Retinopexy
refractive indices. Despite these problems, the procedure was easier to perform under silicone oil as compared to intraocular air. Clinically we observed a narrow white halo of apparent retinal edema around the cyanoacrylate retinopexy site in two of the seven eyes that were enucleated immediately after retinopexy. No distant ocular effects from the procedure could be identified. The average force required to peel the retina from the choroid immediately after surgery was 223.8 mg (range, 150 to 312 mg) at the seven retinotomy sites treated with cyanoacrylate retinopexy and 39.7 mg (range, 16 to 60 mg) at the seven retinotomy sites treated with cryopexy (P<.OOI). In three of the seven cyanoacrylate-treated retinotomies the retina tore at the site of the chorioretinal adhesion during peeling, indicating that the strength of the adhesion produced by the tissue adhesive was greater than the tensile strength of the retina. Of the 19 eyes that were followed up for one week, 16 developed a rim of retinal whitening around the cyanoacrylate retinopexy site. This edema was noticed on the first postoperative day, and it generally started to fade two to three days later. As it faded, a variably severe pigmentary disturbance developed at the level of the retinal pigment epithelium. The size of the retinal whitening was usually small, occurring concentrically, and around the retinotomy site. Occasionally this peculiar retinal whiten-
719
ing was large and eccentrically distributed around the retinotomy site (Fig. 1). The extent of the retinal whitening adjacent to the cyanoacrylate sites appeared to be related to the amount of tissue adhesive that was present, with larger and more severe retinal changes seen with larger quantities of glue. Seven eyes were enucleated one week after the cyanoacrylate retinopexy procedure. The average force needed to peel the retina from the retinal pigment epithelium was 474 mg (range, 164 to 1,118 mg) at the seven retinotomy sites treated with cyanoacrylate retinopexy and 187.4 mg (range, 115 to 322 mg) at the seven retinotomy sites treated with cyropexy (P=.002). The retina tore at the retinotomy site during retinal peeling in four of the cyanoacrylate treated retinotomies and in one of the cryopexy treated retinotomies. Between weeks 1 and 3 there was a further maturation of the pigmentary changes noted around the cyanoacrylate sites. The appearance of this pigmentary reaction was similar to the developing cryopexy scars. When extensive retinal edema subsided massive retinal atrophy developed with subsequent pigment proliferation (Fig. 1). In two cases the retina adjacent to the cyanoacrylate site was so affected that atrophic, crescentic retinal holes developed as the retinal whitening subsided (Fig. 2). With the exception of the two cases in which retinal breaks developed at the edge of the cyanoacry-
Fig. 1 (Sheta, Hida, and McCuen). Left, Fundus photograph of a rabbit eye showing the glue site (arrow) with extensive retinal whitening eccentrically distributed around the glued hole. Right, Fundus photograph of the same eye two weeks later. An area of massive retinal atrophy is seen adjacent to the site of cyanoacrylate retinopexy (arrow).
720
December, 1986
AMERICAN JOURNAL OF OPffiHALMOLOGY
Fig. 2 (Sheta, Hida, and McCuen). Fundus photograph of a rabbit eye showing the site of cyanoacrylate retinopexy (arrow) with a crescentic retinal hole at its edge (arrowhead). Retinal edema is also seen around the glued retinotomy (asterisk). late retinopexy site, peripheral retinal detachments did not occur in this model either intraoperatively or postoperatively. Retinal peeling was performed after four weeks in seven eyes. The eyes that developed retinal tears or massive retinal atrophy at the edge of the cyanoacrylate treated retinotomy were excluded from retinal peeling. The average peeling force required to separate the retina was 582.4 mg (range, 301 to 1,270 mg) at the seven retinotomy sites treated with cyanoacrylate retinopexy and 464.7 mg (range, 181 to 1,118 mg) at the seven retinotomy sites treated with cryopexy. The retina tore during peeling at six of the cyanoacrylate treated retinotomies and at three of the cryopexy treated retinotomies. Although the adhesive force of cyanoacrylate retinopexy was still higher than cryopexy, this was not statistically significant (P=.18). Histologic examination of the retina of one eye 24 hours after cyanoacrylate retinopexy showed clinically evident retinal whitening adjacent to the retinotomy site. Retinal sections studied from the areas of retinal whitening showed a nonspecific coagulative necrosis. There was fragmentation of the inner retina, disorganization of the nuclear layers, and loss of photoreceptors (Fig. 3). The eye receiving pure n-butyl-2-cyanoacrylate without any added iophendylate also developed surround-
Fig. 3 (Sheta, Hida, and McCuen). Histologic section of the retina 24 hours after cyanoacrylate retinopexy. On the right side of the photograph, cyanoacrylate is seen overlying the retinotomy site (arrowheads). Adjacent to the site of cyanoacrylate retinopexy, there is necrosis of the inner retina, disorganization of the nuclear layers, and loss of photoreceptors (arrow). On the left side of the photograph, the retina is normal (hematoxylin and eosin, x 100).
ing retinal whitening similar to most of the other eyes in this study.
Discussion We have previously shown that transvitreal cyanoacrylate retinopexy is technically possible and potentially superior to transscleral retinal cryopexy in the air-filled rabbit and primate eyes.' With the retinal gluing, however, good visualization of the retinal break or retinotomy is necessary to ensure the accurate delivery of tiny amounts of tissue adhesive to the defect. In the management of complex retinal detachments good visualization in the air-filled eye is not always possible, and such a compromised view to the fundus limits the potential applicability of cyanoacrylate retinopexy in these cases. There has been increasing interest in and experience with the use of silicone oil in the management of complex vitreoretinal disorders.v" Silicone oil generally affords excellent intraoperative visualization of the fundus and, at the same time, does not induce polymerization of cyanoacrylate polymers. These advan-
Transvitreal Cyanoacrylate Retinopexy
Vol. 102, No.6
tages of silicone oil over gas stimulated us to evaluate the potential of cyanoacrylate retinopexy in silicone oil-filled eyes. This study has shown that cyanoacrylate retinopexy through silicone oil is technically possible, and that the chorioretinal adhesion produced with cyanoacrylate tissue adhesive is significantly stronger than the adhesion produced by trans scleral retinal cryopexy immediately and one week after application (Fig. 4). Retinal peeling measurements four weeks after cyanoacrylate retinopexy confirmed that the strong chorioretinal adhesions produced by the tissue adhesive are persistent. As was noted with cyanoacrylate retinopexy through air, the strength of the chorioretinal adhesions at the glue sites are stronger at four weeks than they are in the immediate postoperative period. Unexpected and remarkable changes near the cyanoacrylate retinopexy site were found in this model. The whitish halo of apparent retinal edema surrounding the tissue adhesive suggests local tissue toxicity caused by the tissue adhesive. We do not know the exact cause of this exaggerated tissue response to the cyanoacrylate under silicone oil, but it is probably related to either chemical or thermal retinal injury at the time of cyanoacrylate polymerization. The toxic effects of cyanoacrylates are related
7
721
in part to their chemical breakdown products (formaldehyde and cyanoacetate) and to the rate at which they are released. 9.10 The observed changes noted in our model may have been caused by formaldehyde release or alternatively, by the release of impurities in the tissue adhesive preparation at the time of polymerization.":" The occurrence of retinal edema in the single eye in which we used pure cyanoacrylate excludes the iophendylate as the cause of the observed retinal toxicity. Similar retinal and pigment epithelial changes were not observed after cyanoacrylate retinopexy using the same tissue adhesive preparations in the air-filled rabbit eye.' We believe that the silicone bubble may prevent the diffusion of potentially toxic chemical contaminants present in the glue that are released during polymerization. The thermal insulating effect of silicone oil may also concentrate the heat built up at the time of polymerization to the preretinal fluid space. The clinical appearance of the whitish retinal edema and the histologic picture of coagulative necrosis favor a thermal effect, but such an effect does not account for the peculiar shape of retinal edema seen in some cases. We noticed that the smaller the amount of tissue adhesive used, the less the tissue reaction observed (Fig. 5). Excessive glue application does not increase the adhesive strength of a cyanoacrylate bond, and great care must be taken to use only minimal amounts of
6 cu
u ~
.E
5
o-Cyanoacrylate e-Cryopexy
3 -1'"
I
IMMED IWK
4WKS
TIME Fig. 4 (Sheta, Hida, and McCuen). The strength of the chorioretinal adhesion (± two standard deviations) of cyanoacrylate retinopexy compared to transsc1eral retinal cryopexy at various intervals after experimental retinal detachment and repair.
Fig. 5 (Sheta, Hida, and McCuen). Fundus photograph showing that with small amounts of cyanoacrylate there is only slight adjacent tissue reaction.
722
AMERICAN JOURNAL OF OPHTHALMOLOGY
tissue adhesive in clinical situations. Based on the results of our study, we believe that although experimental cyanoacrylate retinopexy through silicone oil has some advantages over the same procedure performed through air, the evident retinal toxicity associated with silicone oil limits its potential clinical usefulness at this time.
References 1. McCuen, B. W., II, Hida, T., Sheta, S. M., Isbey, E. K, Hahn, D. K, and Hickingbotham, D.: Experimental transvitreal cyanoacrylate retinopexy. Am. J. Ophthalmol. 102:199-207, 1986. 2. McCuen, B. W., II, and Hickingbotham, D.: A fiberoptic diathermy tissue manipulator for use in vitreous surgery. Am. J. Ophthalmol. 98:803, 1984. 3. Winer, B. J.: Statistical analysis in experimental design. New York, McGraw-Hill, 1962, pp. 298-318. 4. Grey, R. H. B., and Leaver, P. K: Silicone oil in the treatment of massive preretinal retraction. 1. Results in 105 eyes. Br. J. Ophthalmol. 63:355, 1979. 5. Zivojnovic, R, Mertens, D. A. E., and Peperkamp, E.: Das flussige Silikon in der amotiochirurgie. II. Bericht tiber 280 falle, weitere entwicklung der technik. Klin. Monatsbl. Augenheilkd. 181:444, 1982.
December, 1986
6. Gonvers, M.: Temporary silicone oil tamponade in the management of retinal detachment with proliferative vitreoretinopathy. Am. J. Ophthalmol. 100:239, 1985. 7. McCuen, B. W., II, Landers, M. B., and Machemer, R: The use of silicone oil following failed vitrectomy for retinal detachment with advanced proliferative vitreoretinopathy. Ophthalmology 92:1029, 1985. 8. McCuen, B. W., II, de Juan, E., and Machemer, R.: Silicone oil in vitreoretinal surgery. Part 1. Surgical techniques. Retina 5:189, 1985. 9. Spitznas, M., Lossagk, H., Vogel, M., and [oussen, F.: Intraocular histocompatibility and adhesive strength of butyl-2-cyanoacrylate. Graefes Arch. Klin. Exp. Ophthalmol. 187:102, 1973. 10. Kurokawa, K: Experimental studies on the retinopexy using cyanoacrylate in rabbit's eye. Part 2. Clinical observation. Acta Soc. Ophthalmol. [pn. 76:831, 1972. 11. Leonard, F., Kulkarni, R K, Brandes, G., Nelson, J., and Cameron, J. J.: Synthesis and degradation of poly alkyl alpha-cyanoacrylates. J. Appl. Polymer Sci. 10:259, 1966. 12. Newburn, A. B., and Ziniti, P.: Cell culture toxicity of two cyanoacrylate adhesives. Invest. OphthaI mol. 8:648, 1969. 13. Refojo, M. F., Dohlman, C. H., and Kolispoulas, J.: Adhesives in ophthalmology. A review. Surv. Ophthalmol. 15:217, 1971.
We evaluated the use of transvitreal cyanoacrylate retinopexy in the treatment of experimental rhegmatogenous retinal detachment during vitreous surgery in rabbit eyes filled with silicone oil. The view to the fundus was superior to that obtained in our previous model of cyanoacrylate retinopexy in the airfilled eye. Glue delivery was consequently both easier and more precise through silicone oil relative to air. The chorioretinal adhesions produced with cyanoacrylate tissue adhesive were compared with those produced by transscleral retinal cryopexy and were found to be more rapid in onset as well as stronger. An exaggerated tissue response adjacent to the cyanoacrylate site suggested a potential toxic chemical or thermal reaction, or both, to the tissue adhesive, but there was no evidence of any distant ocular effects. WE HAVE PREVIOUSLY described a model of rhegmatogenous retinal detachment during vitreous surgery in the rabbit with intraoperative repair by fluid-air exchange and the transvitreal application of n-butyl-cyanoacrylate tissue adhesive through the air-filled eye to the retinal break. 1 The chorioretinal adhesions achieved were immediate in onset, stronger than those produced with transscleral retinal cryopexy, and long-lasting in effect. In this study we have modified our experimental model so that cyanoacrylate retinopexy is performed through silicone oil-filled eyes rather than in eyes filled with air.
Accepted for publication Sept. 22, 1986. From the Department of Ophthalmology, Duke University Eye Center, Durham, North Carolina. This study was supported by the Adler Foundation, the National Society to Prevent Blindness, and the Peace Fellowship Program. Reprint requests to Brooks W. McCuen II, M.D., Box 3802, Duke University Eye Center, Durham, NC 27710.
©AMERICAN JOURNAL OF OPHTHALMOLOGY
102:717-722,
Material and Methods Twenty-eight pigmented rabbits weighing from 2.5 to 3.5 kg were used in this study. Two weeks before vitrectomy the planned sclerotomy sites 2 mm posterior to the corneoscleral limbus in the superotemporal, superonasal, and inferotemporal quadrants were marked with 6-0 silk sutures and treated with prophylactic transscleral retinal cryopexy. Three days before vitreous surgery, by indirect ophthalmoscopy, 0.2 cc of pure perfluoropropane gas was injected into the mid-vitreous cavity with a 30-gauge needle through the inferior cryopexy site. The animal was placed upright to prevent the gas from escaping through the injection site. After the intraocular pressure fell below 20 mm Hg, the eyes were reinjected with an additional 0.2 cc of perfluoropropane. One hour before vitreous surgery pupillary dilatation was achieved with topical 0.5% phenylephrine hydrochloride, 0.25% tropicamide, and 1% atropine sulfate. The animals were anesthetized with intramuscular injection of 1.0 ml of ketamine hydrochloride (100 mg/ml) and 0.1 ml of xylazine (100 mg/ml). After retrobulbar injection of 1.0 ml of 1% lidocaine hydrochloride, a conjunctival peritomy was performed for 360 degrees at the corneoscleral limbus. The previously marked sclerotomy sites were identified and a preplaced 7-0 Dexon mattress suture was passed around the inferotemporal site. A 1.4-mm microvitreoretinal blade was used to enter the vitreous cavity at the inferotemporal site and a 2.5-mm infusion cannula delivering Ringer's lactate solution with 5% dextrose was secured in position with the preplaced mattress suture. Similar sclerotomies were performed in the superotemporal and superonasal quadrants. By initiating infusion, a total gas-fluid exchange with refilling of the eye with Ringer's lactate was achieved. A bent 19-9auge butterfly infu-
DECEMBER,
1986
717
718
December, 1986
AMERICAN JOURNAL OF OPHTHALMOLOGY
sion needle was passed through the superonasal sclerotomy while the Shock phacofragmentor needle was inserted through the superotemporal sclerotomy to fragment and aspirate the lens nucleus. The lens cortex and the anterior and posterior lens capsules were removed with the I-mm vitreous cutter. In five cases we left the anterior capsule intact to prevent silicone oil from filling the anterior chamber during later fluid-silicone exchange. After replacing the 19-9auge butterfly needle with the fiberoptic tissue manipulator.! an anterior and posterior vitrectomy was performed with the vitreous cutter. After removing as much cortical vitreous as possible, the vitreous cutter was directed toward the retina, and two retinal holes approximately 500 fL in diameter were created 1 disk diameter above and below the optic disk. An infusion stream of Ringer's lactate from the fiberoptic probe was directed over the retinotomy sites to balloon up the subretinal space and create localized posterior retinal detachments. A fluid-silicone exchange procedure with internal drainage of the subretinal fluid through the iatrogenic retinotomies using a blunt 19-9auge needle then effected retinal reattachment and an eye filled with silicone oil. At this time the retinotomies were randomized so that in each eye one retinotomy was treated with transvitreal cyanoacrylate retinopexy while the other retinotomy was treated with transscleral cryopexy. Cyanoacrylate retinopexy was performed with a 50:50 mixture of n-butyl-2-cyanoacrylate and icphendylate using the specially designed glue applicator previously described." The glue injector was held close to the retinal hole, and a tiny drop of glue was produced at its tip by gentle finger pressure on the actuator. The injector tip with the small bubble of tissue adhesive was brought into contact with the retinotomy site and was then quickly retracted to prevent the injector tip from adhering to the retina during the polymerization of the adhesive. Transscleral retinal cryopexy was performed under microscopic control after closure of the two superior sclerotomies with 7-0 Dexon sutures. A bright white choroidal and retinal freeze was produced to surround the retinotomy while making sure that the cryopexy reaction occurred only at the retinotomy site and that it approximated the size of the cyanoacrylate application (approximately 1,000 u.). The infusion cannula was removed and its scleroto-
my closed with a preplaced mattress suture. Ten milligrams of gentamicin was injected into the subconjunctival space, and 1% atropine and dexamethasone and polymixin B sulfate ointments were applied topically to the globe. Seven eyes were enucleated immediately after surgery and the adhesive forces were measured using retinal peeling at the retinotomy site.! In seven eyes ophthalmoscopic and slit-lamp examinations were performed on days 2, 4, and 7, after which the eyes were enucleated and adhesive force measurements were performed. The 12 eyes in the third group were examined at days 2, 4, 7, and weekly thereafter for one month, at which time the eyes were enucleated and adhesive force measurements were performed. An additional eye was enucleated 24 hours after surgery and fixed in 10% buffered formalin. Selected specimens were stained with hematoxylin and eosin and oil red 0 stains for histologic evaluation. In one other eye we used pure cyanoacrylate without iophendylate to seal the retinal hole. Statistical analysis of the chorioretinal adhesive forces was done after logarithmic transformation of the original measure of force since the coefficient of variation remained constant over a wide range of means. The log transformed data were analyzed using a repeated measures two-way analysis of variance design. The two main factors were the method of retinopexy and the time period of the peeling."
Results We encountered several technical problems associated with cyanoacrylate retinopexy in the silicone oil-filled eyes. During cyanoacrylate delivery the tissue adhesive polymerizes more slowly under silicone oil than under air. Although slower polymerization could lead to undesired spreading of glue onto the adjacent retina, this was prevented by using only small amounts of tissue adhesive. Additionally, the glue has a specific gravity less than that of silicone oil and the cyanoacrylate drop tends to float upward from the retinal surface. This tendency was counteracted by lightly pushing the bubble of glue downward against the floor of the retinal break until polymerization began to occur. Also, visualization of the transparent glue drop was difficult in the silicone-filled eye because of the similarity of their respective
Vol. 102, No.6
Transvitreal Cyanoacrylate Retinopexy
refractive indices. Despite these problems, the procedure was easier to perform under silicone oil as compared to intraocular air. Clinically we observed a narrow white halo of apparent retinal edema around the cyanoacrylate retinopexy site in two of the seven eyes that were enucleated immediately after retinopexy. No distant ocular effects from the procedure could be identified. The average force required to peel the retina from the choroid immediately after surgery was 223.8 mg (range, 150 to 312 mg) at the seven retinotomy sites treated with cyanoacrylate retinopexy and 39.7 mg (range, 16 to 60 mg) at the seven retinotomy sites treated with cryopexy (P<.OOI). In three of the seven cyanoacrylate-treated retinotomies the retina tore at the site of the chorioretinal adhesion during peeling, indicating that the strength of the adhesion produced by the tissue adhesive was greater than the tensile strength of the retina. Of the 19 eyes that were followed up for one week, 16 developed a rim of retinal whitening around the cyanoacrylate retinopexy site. This edema was noticed on the first postoperative day, and it generally started to fade two to three days later. As it faded, a variably severe pigmentary disturbance developed at the level of the retinal pigment epithelium. The size of the retinal whitening was usually small, occurring concentrically, and around the retinotomy site. Occasionally this peculiar retinal whiten-
719
ing was large and eccentrically distributed around the retinotomy site (Fig. 1). The extent of the retinal whitening adjacent to the cyanoacrylate sites appeared to be related to the amount of tissue adhesive that was present, with larger and more severe retinal changes seen with larger quantities of glue. Seven eyes were enucleated one week after the cyanoacrylate retinopexy procedure. The average force needed to peel the retina from the retinal pigment epithelium was 474 mg (range, 164 to 1,118 mg) at the seven retinotomy sites treated with cyanoacrylate retinopexy and 187.4 mg (range, 115 to 322 mg) at the seven retinotomy sites treated with cyropexy (P=.002). The retina tore at the retinotomy site during retinal peeling in four of the cyanoacrylate treated retinotomies and in one of the cryopexy treated retinotomies. Between weeks 1 and 3 there was a further maturation of the pigmentary changes noted around the cyanoacrylate sites. The appearance of this pigmentary reaction was similar to the developing cryopexy scars. When extensive retinal edema subsided massive retinal atrophy developed with subsequent pigment proliferation (Fig. 1). In two cases the retina adjacent to the cyanoacrylate site was so affected that atrophic, crescentic retinal holes developed as the retinal whitening subsided (Fig. 2). With the exception of the two cases in which retinal breaks developed at the edge of the cyanoacry-
Fig. 1 (Sheta, Hida, and McCuen). Left, Fundus photograph of a rabbit eye showing the glue site (arrow) with extensive retinal whitening eccentrically distributed around the glued hole. Right, Fundus photograph of the same eye two weeks later. An area of massive retinal atrophy is seen adjacent to the site of cyanoacrylate retinopexy (arrow).
720
December, 1986
AMERICAN JOURNAL OF OPffiHALMOLOGY
Fig. 2 (Sheta, Hida, and McCuen). Fundus photograph of a rabbit eye showing the site of cyanoacrylate retinopexy (arrow) with a crescentic retinal hole at its edge (arrowhead). Retinal edema is also seen around the glued retinotomy (asterisk). late retinopexy site, peripheral retinal detachments did not occur in this model either intraoperatively or postoperatively. Retinal peeling was performed after four weeks in seven eyes. The eyes that developed retinal tears or massive retinal atrophy at the edge of the cyanoacrylate treated retinotomy were excluded from retinal peeling. The average peeling force required to separate the retina was 582.4 mg (range, 301 to 1,270 mg) at the seven retinotomy sites treated with cyanoacrylate retinopexy and 464.7 mg (range, 181 to 1,118 mg) at the seven retinotomy sites treated with cryopexy. The retina tore during peeling at six of the cyanoacrylate treated retinotomies and at three of the cryopexy treated retinotomies. Although the adhesive force of cyanoacrylate retinopexy was still higher than cryopexy, this was not statistically significant (P=.18). Histologic examination of the retina of one eye 24 hours after cyanoacrylate retinopexy showed clinically evident retinal whitening adjacent to the retinotomy site. Retinal sections studied from the areas of retinal whitening showed a nonspecific coagulative necrosis. There was fragmentation of the inner retina, disorganization of the nuclear layers, and loss of photoreceptors (Fig. 3). The eye receiving pure n-butyl-2-cyanoacrylate without any added iophendylate also developed surround-
Fig. 3 (Sheta, Hida, and McCuen). Histologic section of the retina 24 hours after cyanoacrylate retinopexy. On the right side of the photograph, cyanoacrylate is seen overlying the retinotomy site (arrowheads). Adjacent to the site of cyanoacrylate retinopexy, there is necrosis of the inner retina, disorganization of the nuclear layers, and loss of photoreceptors (arrow). On the left side of the photograph, the retina is normal (hematoxylin and eosin, x 100).
ing retinal whitening similar to most of the other eyes in this study.
Discussion We have previously shown that transvitreal cyanoacrylate retinopexy is technically possible and potentially superior to transscleral retinal cryopexy in the air-filled rabbit and primate eyes.' With the retinal gluing, however, good visualization of the retinal break or retinotomy is necessary to ensure the accurate delivery of tiny amounts of tissue adhesive to the defect. In the management of complex retinal detachments good visualization in the air-filled eye is not always possible, and such a compromised view to the fundus limits the potential applicability of cyanoacrylate retinopexy in these cases. There has been increasing interest in and experience with the use of silicone oil in the management of complex vitreoretinal disorders.v" Silicone oil generally affords excellent intraoperative visualization of the fundus and, at the same time, does not induce polymerization of cyanoacrylate polymers. These advan-
Transvitreal Cyanoacrylate Retinopexy
Vol. 102, No.6
tages of silicone oil over gas stimulated us to evaluate the potential of cyanoacrylate retinopexy in silicone oil-filled eyes. This study has shown that cyanoacrylate retinopexy through silicone oil is technically possible, and that the chorioretinal adhesion produced with cyanoacrylate tissue adhesive is significantly stronger than the adhesion produced by trans scleral retinal cryopexy immediately and one week after application (Fig. 4). Retinal peeling measurements four weeks after cyanoacrylate retinopexy confirmed that the strong chorioretinal adhesions produced by the tissue adhesive are persistent. As was noted with cyanoacrylate retinopexy through air, the strength of the chorioretinal adhesions at the glue sites are stronger at four weeks than they are in the immediate postoperative period. Unexpected and remarkable changes near the cyanoacrylate retinopexy site were found in this model. The whitish halo of apparent retinal edema surrounding the tissue adhesive suggests local tissue toxicity caused by the tissue adhesive. We do not know the exact cause of this exaggerated tissue response to the cyanoacrylate under silicone oil, but it is probably related to either chemical or thermal retinal injury at the time of cyanoacrylate polymerization. The toxic effects of cyanoacrylates are related
7
721
in part to their chemical breakdown products (formaldehyde and cyanoacetate) and to the rate at which they are released. 9.10 The observed changes noted in our model may have been caused by formaldehyde release or alternatively, by the release of impurities in the tissue adhesive preparation at the time of polymerization.":" The occurrence of retinal edema in the single eye in which we used pure cyanoacrylate excludes the iophendylate as the cause of the observed retinal toxicity. Similar retinal and pigment epithelial changes were not observed after cyanoacrylate retinopexy using the same tissue adhesive preparations in the air-filled rabbit eye.' We believe that the silicone bubble may prevent the diffusion of potentially toxic chemical contaminants present in the glue that are released during polymerization. The thermal insulating effect of silicone oil may also concentrate the heat built up at the time of polymerization to the preretinal fluid space. The clinical appearance of the whitish retinal edema and the histologic picture of coagulative necrosis favor a thermal effect, but such an effect does not account for the peculiar shape of retinal edema seen in some cases. We noticed that the smaller the amount of tissue adhesive used, the less the tissue reaction observed (Fig. 5). Excessive glue application does not increase the adhesive strength of a cyanoacrylate bond, and great care must be taken to use only minimal amounts of
6 cu
u ~
.E
5
o-Cyanoacrylate e-Cryopexy
3 -1'"
I
IMMED IWK
4WKS
TIME Fig. 4 (Sheta, Hida, and McCuen). The strength of the chorioretinal adhesion (± two standard deviations) of cyanoacrylate retinopexy compared to transsc1eral retinal cryopexy at various intervals after experimental retinal detachment and repair.
Fig. 5 (Sheta, Hida, and McCuen). Fundus photograph showing that with small amounts of cyanoacrylate there is only slight adjacent tissue reaction.
722
AMERICAN JOURNAL OF OPHTHALMOLOGY
tissue adhesive in clinical situations. Based on the results of our study, we believe that although experimental cyanoacrylate retinopexy through silicone oil has some advantages over the same procedure performed through air, the evident retinal toxicity associated with silicone oil limits its potential clinical usefulness at this time.
References 1. McCuen, B. W., II, Hida, T., Sheta, S. M., Isbey, E. K, Hahn, D. K, and Hickingbotham, D.: Experimental transvitreal cyanoacrylate retinopexy. Am. J. Ophthalmol. 102:199-207, 1986. 2. McCuen, B. W., II, and Hickingbotham, D.: A fiberoptic diathermy tissue manipulator for use in vitreous surgery. Am. J. Ophthalmol. 98:803, 1984. 3. Winer, B. J.: Statistical analysis in experimental design. New York, McGraw-Hill, 1962, pp. 298-318. 4. Grey, R. H. B., and Leaver, P. K: Silicone oil in the treatment of massive preretinal retraction. 1. Results in 105 eyes. Br. J. Ophthalmol. 63:355, 1979. 5. Zivojnovic, R, Mertens, D. A. E., and Peperkamp, E.: Das flussige Silikon in der amotiochirurgie. II. Bericht tiber 280 falle, weitere entwicklung der technik. Klin. Monatsbl. Augenheilkd. 181:444, 1982.
December, 1986
6. Gonvers, M.: Temporary silicone oil tamponade in the management of retinal detachment with proliferative vitreoretinopathy. Am. J. Ophthalmol. 100:239, 1985. 7. McCuen, B. W., II, Landers, M. B., and Machemer, R: The use of silicone oil following failed vitrectomy for retinal detachment with advanced proliferative vitreoretinopathy. Ophthalmology 92:1029, 1985. 8. McCuen, B. W., II, de Juan, E., and Machemer, R.: Silicone oil in vitreoretinal surgery. Part 1. Surgical techniques. Retina 5:189, 1985. 9. Spitznas, M., Lossagk, H., Vogel, M., and [oussen, F.: Intraocular histocompatibility and adhesive strength of butyl-2-cyanoacrylate. Graefes Arch. Klin. Exp. Ophthalmol. 187:102, 1973. 10. Kurokawa, K: Experimental studies on the retinopexy using cyanoacrylate in rabbit's eye. Part 2. Clinical observation. Acta Soc. Ophthalmol. [pn. 76:831, 1972. 11. Leonard, F., Kulkarni, R K, Brandes, G., Nelson, J., and Cameron, J. J.: Synthesis and degradation of poly alkyl alpha-cyanoacrylates. J. Appl. Polymer Sci. 10:259, 1966. 12. Newburn, A. B., and Ziniti, P.: Cell culture toxicity of two cyanoacrylate adhesives. Invest. OphthaI mol. 8:648, 1969. 13. Refojo, M. F., Dohlman, C. H., and Kolispoulas, J.: Adhesives in ophthalmology. A review. Surv. Ophthalmol. 15:217, 1971.