- Email: [email protected]
Electron Microscopic Study of Intrastromal Hydrogel Implants in Primates
Electron Microscopic Study of Intrastromal Hydrogel Implants in Primates TATSUO YAMAGUCHI, MD, STEVEN B. KOENIG, MD,* TAKASHI HAMANO, MD, TAIRO KIMURA, MD, ELIZABETH SANTANA, MD,t MARGUERITE B. McDONALD, MD, HERBERT E. KAUFMAN, MD
Abstract: Three corneas with intrastromal hydrogel implants (surfilcon A) were removed from Green monkey eyes by penetrating keratoplasty and examined by light microscopy, as well as scanning and transmission electron microscopy, in order to assess the tolerance of the primate cornea for this type of synthetic plastic. Placement of these implants in the posterior cornea appears to increase the amount of protrusion into the anterior chamber, which can also be seen clinically on slit-lamp biomicroscopy. Some physiological stress to the cornea was indicated by abnormalities above and below the implant, including thinned epithelium, irregular cell shapes, and vacuolations. It appears that a basement-membrane-like material is produced by keratocytes adjacent to the implant-stroma interface, and that this material fills the spaces and provides some physical continuity between the plastic and the corneal tissue. No inflammatory reaction was seen around the implants, but further long-term studies are necessary to ensure compatibility between the cornea and the implant. [Key words: biocompatibility, cornea, electron microscopy, hydrogel implants, intrastromal implants.] Ophthalmology 91 :1170-1175, 1984
Intrastromal implants made of plastic materials have been used in the past to investigate the physiology of the cornea and for the therapy of epithelial and endothelial dystrophies. I - 7 A number of different kinds of plastic have been tried, such as polyethylene, acryl (methacrylate tetron), polyvinylidine, polypropylene, and silicon, among others. However, experimental studies have been unable to document long-term corneal tolerance for these synthetic substances; therefore, their use as permanent additions to the eye has heretofore not been possible. From the Lions Eye Research Laboratories, LSU Eye Center, Louisiana State University Medical Center School of Medicine, New Orleans. *Or. Koenig is currently located at the Medical College of Wisconsin, Milwaukee and tOr. Santana is located in Caracas, Venezuela. Supported in part by USPHS grants EY04082, EY03635, EY02580, and EY02377 from the National Eye Institute and a Core grant (5-P40RR00164) to the Delta Regional Primate Research Center, Covington, LouiSiana, from the National Institutes of Health, Bethesda, Maryland. Reprint requests to Tatsuo Yamaguchi, MO, LSU Eye Center, 136 South Roman St., New Orleans, LA 70112.
1170
With the advent of the new, high-water-content, hydrogel plastics,8-10 there has been a renewed interest in this subject, concurrent with the recent upsurge in interest in surgical procedures for the correction of refractive error. 11 - 16 Present limitations of the various refractive techniques include lack of precision in the final dioptric outcome, inadequate dioptric power as a result of insufficient thickness in the donor tissue, and a chronic shortage of donor tissue. Some of these problems could be alleviated by means of techniques incorporating plastic materials shaped to specific powers, provided that such an implant can be tolerated by the cornea. Intrastromal hydrogel lenses have been implanted experimentally in rabbit and monkey eyes,8-16 and in one human subject,17 with varying degrees of success in terms of refractive increase and corneal tolerance, both dependent to some extent on the surgical procedure employed. We report the light and electron microscopic findings in three monkey corneas that had received intrastromal implants nine to ten months previously, in order to assess the anatomic and histologic effects of
YAMAGUCHI, et al
•
HYDROGEL IMPLANTS IN PRIMATES
I
Implant
AC
*
Fig 1. Light micrograph. Right, center; left, periphery. Hydrogel lens implant (implant) placed at 4/5 the depth of the cornea protrudes into the anterior chamber (AC). Note that the epithelial cell layer is thinner in the center (arrow) over the implant and thicker in the periphery. Ant. St = anterior stroma; .. indicates posterior stroma (Toluidine blue, X1l4).
this plastic material in eyes that clinically appeared to be successfully tolerating the presence of the foreign material.
MATERIALS AND METHODS One eye from each of three Green monkeys (Ceropithecus aethiops) received an intrastromal implant composed of Duragel 75® (surfilcon A; CooperVision, San Jose, California), which consists of a hydrophobic backbone of polymethylmethacrylate complexed with hydrophilic poly-N-2-vinyl pyrrolidone. The plastic has the following characteristics: 73.5% water content at 25°C; 43% oxygen transmissibility at 20°C, and 45% at 35°C; 23 A pore size at 25°C; and 1.37 index of refraction. All implants were lathe cut to the following specifications: 5.4 mm base curve; +14.00 diopter power; 5.0 mm diameter; and 0.2 mm central thickness. For the implantation surgery, the monkeys were anesthetized with halothane. The intrastromal pocket was created by means of a Paufique knife and cyclodialysis spatula through a 60° to 90° incision. The implant was inserted and the incision was closed with a running 10o nylon suture. Topical atropine 1% and gentamicin ointments were applied to the cornea. Postoperative examinations were performed weekly during the early postoperative period. Sutures were removed six weeks after surgery. Keratometry and pachymetry measurements were recorded at one, two, three, and eight months postoperatively. Corneal photography was performed with a Nikon slit-lamp camera. Nine and a half and ten months after implantation surgery, the corneas were removed by penetrating keratoplasty. The excised corneal buttons containing the implants were fixed immediately in 2.5% glutaraldehyde
and 3% formalin with 1/15 M potassium-sodium phosphate buffer (pH 7.4; final osmolarity of fixative and buffer-360 mOsm) for 90 minutes, washed in the same buffer, and dissected. For light microscopy (LM) and transmission electron microscopy (TEM), the specimens were postfixed in 1% osmium tetroxide with the same buffer for 90 minutes, washed again in buffer, dehydrated in graded alcohols from 50% to 100% and in propylene oxide, and embedded in Epon. Embedded specimens were sectioned with a microtome and stained with toluidine blue for LM or with uranyl acetate and lead citrate for TEM. For scanning electron microscopy (SEM), after dehydration, the specimens were dried in a critical-point dryer, placed on aluminum stubs, and coated with gold.
RESULTS LIGHT MICROSCOPY
Despite the attempt to locate the intrastromal pockets at approximately half the depth of the cornea, light microscopic histologic findings showed that the hydrogel lenticules were actually situated from 3/4 to 4/5 of the stromal depth (Fig 1). The deeper the implant, the more the posterior cornea was seen to protrude into the anterior chamber. The protrusion was also seen on slitlamp examination in the monkey eyes, but was more pronounced in the histologic specimens. The epithelial cell layer in all the specimens was thinner in the central area over the implant, compared to the periphery outside the limits of the pocket and lenticule. In the central area, the epithelial cells were flat and irregular in shape. Bowman's membrane and the anterior stroma looked 1171
OPHTHALMOLOGY
•
OCTOBER 1984 •
VOLUME 91
•
NUMBER 10
Implant
AC Fig 2. Light micrograph. In the area of the protrusion beneath the edge of the hydrogel lens implant (implant), there is a fold in Descemet's membrane and some endothelial vacuolations can be seen (small arrows). Between the edges of the implant and the stroma, newly formed collagen fibrils are seen (large arrow). AC = anterior chamber (Toluidine blue, X287).
Fig 3. Epithelial cells in the central cornea are irregular in shape and have many vacuolations (V). BM = Bowman's membrane. (TEM, X7llO; bar = 2.8 JLm).
normal. No inflammatory cells were seen in the vicinity of the implant. Between the edges of the implant and the stroma, newly formed collagen fibrils and some keratocytes were seen. The posterior cornea appeared normal. Vacuolations were seen more frequently in some endothelial cells in the central area, compared to the periphery. Beneath the edges of the implant, where the posterior corneal protrusion into the anterior chamber began, a fold in Descemet's membrane and some vacuolations in the endothelial cells were seen (Fig 2). ELECTRON MICROSCOPY
The epithelial cells in the central cornea were irregular in shape and contained many vacuolations (Fig 3). The epithelial cells in the periphery looked normal, except for some edematous rough endoplasmic recticulum (Fig 4). In the anterior stroma over the implant, the collagen lamellae were slightly irregular and the cytoplasm in some keratocytes appeared to be of non-uniform density, indicating the early stages of degeneration (Fig. 5). At the outer edges of the intrastromal pocket, where the implant did not entirely fill the space made by the spatula at the time of surgery, fibrinous tissue in an amorphous matrix and some degenerated keratocytes were seen to form a connecting layer about 105 to 180 nm thick between the anterior and the posterior cornea (Fig 6). A similar material was seen between the implant and some stromal keratocytes immediately adjacent to the implant both anteriorly and posteriorly (Fig 7). Relatively more amorphous matrix was found between implant and keratocyte; more fibrinous tissue (55-140 A) was found between keratocyte and stroma. These keratocytes were sparsely distributed both anteriorly and posteriorly in the central area, but increased in numbers toward the periphery, where some accumulation of these lin
Fig 4. In the periphery, the epithelial basal cells look normal except for edematous rough endoplasmic reticulum (small arrows). The large arrow indicates a ferritin deposit in the intracellular space. Note the same material in the cy1oplasm. BM = Bowman's membrane; N = nucleus. (TEM, X7llO; bar = 2.8 JLm).
cells was seen. Macrophage and fibroblastic cells were observed between the implant and the stroma, also both anteriorly and posteriorly (Fig 8). The posterior cornea beneath the implant looked normal. Centrally, the endothelial cells were flat and showed some cytoplasmic vacuolations (Fig 9). Newly formed abnormal collagen fibrils were seen in the posterior Descemet's membrane. Scanning electron microscopy of the endothelial cells in the central area showed some edematous cells, as well as some with depressions in the middle and at the cell borders (Fig 10). The peripheral endothelial cells looked more normal by both transmission (Fig II) and scanning (Fig 12) electron microscopy, except for slight depression of the cell borders (Fig 12). No abnormal collagen fibrils were seen in the posterior Descemet's membrane peripherally.
YAMAGUCHI , et al
• HYDROGEL IMPLANTS IN PRIMATES
Fig S. Keratocyte (K) in anterior cornea over the implant. Nonuniform cytoplasmic density indicates early stages of degeneration. (TEM, X 10,428; bar = 1.9 I'm).
Fig 7. Interface between the posterior cornea (Post. St) and the hydrogel lens (implant). Newly formed fibrinous tissue in basement membranelike amorphous matrix (arrows) can be seen between the hydrogel lens and a keratocyte (K) and between a keratocyte and the posterior cornea; * indicates separation from posterior cornea during dissection (TEM, X47400, bar = 0.4 I'm).
DISCUSSION McCarey and Andrews l2 implanted 71 % water content lenses in rabbit corneas and found slowly progressive thinning of the stroma anterior to the edges of the impla.nt. They suggested that this complication was a result of mechanical stress. However, in our study, more abnormalities appeared in the central areas above and below the implant, and the character of these abnormalities indicated some type of physiological stress. The epithelial layer was thinner and the cells were more irregular in shape and had more vacuo lations over the implant, compared to the periphery (Figs 3, 4). Hamano et al. 10 also reported abnormalities of the corneal epithelium over intrastromal hydrogel implants observed by scanning electron microscopy. Also, we observed
Fig 6. Junction between anterior cornea (Ant. St) and posterior cornea (Post. St) at edge of intrastromal pocket. The wound, which was made by the spatula, is filled with amorphous matrix (arrows). A degenerated keratocyte (K) is also seen. (TEM, X28440; bar = 0.7 I'm).
Fig 8. Area between the hydrogel lens implant (implant), anterior cornea (Ant. St), and posterior cornea (Post. St); * indicates artifactitious separation during dissection. Note macrophage (large arrow) and fibroblasts (small arrow). (TEM, X711O; bar = 2.8 I'm).
abnormal collagen fibrils, presumably secreted by the vacuolated endothelial cells, on Descemet's membrane (Fig 9). Furthermore, some keratocytes immediately anterior to the implant displayed degenerative changes (Fig 5). In contrast, the peripheral areas of the cornea were more normal looking. It may be that the implant produces some endothelial changes, as well as impeding the nutrient flow from the aqueous humor toward the epithelium, resulting in the observed abnormalities. Although the posterior protrusion into the anterior chamber appeared to be more pronounced with more posterior placement of these implants, a more anterior placement would seem likely to carry an increased risk of anterior perforation or extrusion. Inasmuch as the procedure used here did not include a microkeratome lamellar dissection, no refractive increase was obtained, because of the barrier nature of Bowman's membrane and the anterior collagen lamellae, which prevents an1173
OPHTHALMOLOGY
•
OCTOBER 1984
•
VOLUME 91
•
NUMBER 10
AC Fig 9. Vacuoles (V) can be seen in the endothelial cells in the central cornea beneath the implant. Note newly formed abnormal collagen fibrils (arrows) on posterior Oescemet's membrane (OM). N = nucleus; m = mitochondria; AC = anterior chamber. (TEM, X16590; bar =
Fig 10. In the central cornea beneath the implant, the middles and cell borders of the endothelial cells are depressed (arrows). (SEM, X2152; bar = 4.6 I'm).
1.2 I'm).
AC Fig 11. In the periphery beyond the edges of the implant, the endothelial cells are more normal looking than the cells in the central area (Fig 9). The cristae of the mitochondria (m) are not clear. No abnormal collagen fibrils can be seen in the posterior Descement's membrane. OM = Descemet's membrane; AC = anterior chamber; N = nucleus. (TEM, X16590; bar = 1.2 I'm).
terior bowing of the cornea. However, the question of placement depth may apply to the keratophakia procedure as well, and a 50% depth has been suggested to be optimal by Werblin et al. 16 Our experience in this study was that although the pockets were created at 50% depth the implants were actually located from 3/4 to 4/5 the depth of the cornea. This problem clearly requires further work to ascertain the importance of implant depth and greater reliability of placement. In our study, keratocytes were seen in the stroma adjacent to the implant, distributed somewhat thinly centrally and more densely in the periphery. Also, keratocytes and newly formed fibrinous tissue in a basement-membrane-like amorphous matrix were ob1174
Fig 12. Endothelial cells look normal in the periphery except for the depressions at the cell boundaries (arrows). (SEM, X2182; bar = 4.6 I'm).
served between the implant and the stroma. Sendele et al 17 reported a case of hydrogel implantation in a human eye. Collagenous filaments and lined fibroblasts were seen at the implant/stroma interface six months after implantation, but no basement-membrane-like material was observed in the space between the plastic and the cornea. However, there appeared to be an artifactitious separation between the implant and the tissue, so such material may have been lost in the processing of the specimens. In any case, it appears that in both primate and human eyes, keratocytes play some role in secreting material to fill the spaces between the tissue and the plastic. No inflammatory reaction was seen around the implant, either in primate eyes in our study, or in the human eye examined by Sendele et al. 17 Plastic implants have considerable potential for mitigating some of the problems connected with current refractive surgical pro-
YAMAGUCHI, et al
•
HYDROGEL IMPLANTS IN PRIMATES
cedures, but further investigation of corneal tolerance for these synthetic materials will be needed to ensure long-term safety.
ACKNOWLEDGMENT The authors thank Mr. Hewitt Cabirac, of the Delta Regional Primate Research Center, Covington, Louisiana, for his assistance with the monkeys used in this study.
REFERENCES 1. Bock RH, Maumenee AE. Corneal fluid metabolism; experiments and observations. Arch Ophthalmol 1953; 50:282-5. 2. Stone W Jr, Herbert E. Experimental study of plastic material as replacement for the cornea; a preliminary report. Am J Ophthalmol 1953; 36(P!. 2): 168-73. 3. Krwawicz T. New plastic operation for correcting the refractive error of aphakic eyes by changing the corneal curvature; preliminary report. Br J Ophthalmol 1961; 45:59-63. 4. Knowles WF. Effect of intralamellar plastic membranes on corneal physiology. Am J Ophthalmol 1961; 51 :1146-56. 5. Belau PG, Dyer JA, Ogle KN, Henderson Jw. Correction of ametropia with intracorneal lenses; an experimental study. Arch Ophthalmol 1964; 72:541-7.
6. Onuma M, Hayashi T. Intralamellar transplantation of acrylic meshes and discs. Jpn J Clin Ophthalmol 1963; 17:453-7. 7. Brown SI, Dohlman CH. A buried corneal implant serving as a barrier to fluid. Arch Ophthalmol 1965; 73:635-9. 8. Dohlman CH, Refojo MF, Rose J. Synthetic polymers in corneal surgery. I. Glyceryl methacrylate. Arch Ophthalmol 167; 77:252-7. 9. Otsuka J, Tokoro T, Tarumi N, et al. On the influences hydrogel membranes on the cornea of rabbits. J Jpn Contact Lens Soc 1971; 13:140-4. 10. Hamano H, Hori M, Hirayama K. Scanning electron microscope observation of corneal epithelium. Part VI. Intralamellar implant of hydrogel contact lens. J Jpn Contact Lens Soc 1973; 15:111-22. 11. Mester U, Heimig 0, Dardenne MU. Measurement and calculation of refraction in experimental keratophakia with hydrophilic lenses. Ophthalmic Res 1976; 8:111-6. 12. McCarey BE, Andrews OM. Refractive keratoplasty with intrastromal hydrogel lenticular implants. Invest Ophthalmol Vis Sci 1981; 21 :10715. 13. Binder PS, Deg JK, Zavala EY, Grossman KR. Hydrogel keratophakia in non-human primates. Curr Eye Res 1982; 1:535-42. 14. Sher JH, Dixon WS. Interlamellar refractive keratoplasty in rabbits. Can J Ophthalmol1982; 17:116-20. 15. McDonald MB, Koenig SB, Friedlander MH, et al. Alloplastic epikeratophakia for the correction of aphakia. Ophthalmic Surg 1983; 14:65-9. 16. Werblin TP, Blaydes JE, Fryczkowski AW, Peiffer R. Stability of hydrogel intracorneal implants in non-human primates. CLAO J 1983; 9:157-61. 17. Sendele DO, Abelson MB, Kenyon KR, Hanninen LA. Intracorneal lens implantation. Arch Ophthalmol 1983; 101 :940-4.
1175
Abstract: Three corneas with intrastromal hydrogel implants (surfilcon A) were removed from Green monkey eyes by penetrating keratoplasty and examined by light microscopy, as well as scanning and transmission electron microscopy, in order to assess the tolerance of the primate cornea for this type of synthetic plastic. Placement of these implants in the posterior cornea appears to increase the amount of protrusion into the anterior chamber, which can also be seen clinically on slit-lamp biomicroscopy. Some physiological stress to the cornea was indicated by abnormalities above and below the implant, including thinned epithelium, irregular cell shapes, and vacuolations. It appears that a basement-membrane-like material is produced by keratocytes adjacent to the implant-stroma interface, and that this material fills the spaces and provides some physical continuity between the plastic and the corneal tissue. No inflammatory reaction was seen around the implants, but further long-term studies are necessary to ensure compatibility between the cornea and the implant. [Key words: biocompatibility, cornea, electron microscopy, hydrogel implants, intrastromal implants.] Ophthalmology 91 :1170-1175, 1984
Intrastromal implants made of plastic materials have been used in the past to investigate the physiology of the cornea and for the therapy of epithelial and endothelial dystrophies. I - 7 A number of different kinds of plastic have been tried, such as polyethylene, acryl (methacrylate tetron), polyvinylidine, polypropylene, and silicon, among others. However, experimental studies have been unable to document long-term corneal tolerance for these synthetic substances; therefore, their use as permanent additions to the eye has heretofore not been possible. From the Lions Eye Research Laboratories, LSU Eye Center, Louisiana State University Medical Center School of Medicine, New Orleans. *Or. Koenig is currently located at the Medical College of Wisconsin, Milwaukee and tOr. Santana is located in Caracas, Venezuela. Supported in part by USPHS grants EY04082, EY03635, EY02580, and EY02377 from the National Eye Institute and a Core grant (5-P40RR00164) to the Delta Regional Primate Research Center, Covington, LouiSiana, from the National Institutes of Health, Bethesda, Maryland. Reprint requests to Tatsuo Yamaguchi, MO, LSU Eye Center, 136 South Roman St., New Orleans, LA 70112.
1170
With the advent of the new, high-water-content, hydrogel plastics,8-10 there has been a renewed interest in this subject, concurrent with the recent upsurge in interest in surgical procedures for the correction of refractive error. 11 - 16 Present limitations of the various refractive techniques include lack of precision in the final dioptric outcome, inadequate dioptric power as a result of insufficient thickness in the donor tissue, and a chronic shortage of donor tissue. Some of these problems could be alleviated by means of techniques incorporating plastic materials shaped to specific powers, provided that such an implant can be tolerated by the cornea. Intrastromal hydrogel lenses have been implanted experimentally in rabbit and monkey eyes,8-16 and in one human subject,17 with varying degrees of success in terms of refractive increase and corneal tolerance, both dependent to some extent on the surgical procedure employed. We report the light and electron microscopic findings in three monkey corneas that had received intrastromal implants nine to ten months previously, in order to assess the anatomic and histologic effects of
YAMAGUCHI, et al
•
HYDROGEL IMPLANTS IN PRIMATES
I
Implant
AC
*
Fig 1. Light micrograph. Right, center; left, periphery. Hydrogel lens implant (implant) placed at 4/5 the depth of the cornea protrudes into the anterior chamber (AC). Note that the epithelial cell layer is thinner in the center (arrow) over the implant and thicker in the periphery. Ant. St = anterior stroma; .. indicates posterior stroma (Toluidine blue, X1l4).
this plastic material in eyes that clinically appeared to be successfully tolerating the presence of the foreign material.
MATERIALS AND METHODS One eye from each of three Green monkeys (Ceropithecus aethiops) received an intrastromal implant composed of Duragel 75® (surfilcon A; CooperVision, San Jose, California), which consists of a hydrophobic backbone of polymethylmethacrylate complexed with hydrophilic poly-N-2-vinyl pyrrolidone. The plastic has the following characteristics: 73.5% water content at 25°C; 43% oxygen transmissibility at 20°C, and 45% at 35°C; 23 A pore size at 25°C; and 1.37 index of refraction. All implants were lathe cut to the following specifications: 5.4 mm base curve; +14.00 diopter power; 5.0 mm diameter; and 0.2 mm central thickness. For the implantation surgery, the monkeys were anesthetized with halothane. The intrastromal pocket was created by means of a Paufique knife and cyclodialysis spatula through a 60° to 90° incision. The implant was inserted and the incision was closed with a running 10o nylon suture. Topical atropine 1% and gentamicin ointments were applied to the cornea. Postoperative examinations were performed weekly during the early postoperative period. Sutures were removed six weeks after surgery. Keratometry and pachymetry measurements were recorded at one, two, three, and eight months postoperatively. Corneal photography was performed with a Nikon slit-lamp camera. Nine and a half and ten months after implantation surgery, the corneas were removed by penetrating keratoplasty. The excised corneal buttons containing the implants were fixed immediately in 2.5% glutaraldehyde
and 3% formalin with 1/15 M potassium-sodium phosphate buffer (pH 7.4; final osmolarity of fixative and buffer-360 mOsm) for 90 minutes, washed in the same buffer, and dissected. For light microscopy (LM) and transmission electron microscopy (TEM), the specimens were postfixed in 1% osmium tetroxide with the same buffer for 90 minutes, washed again in buffer, dehydrated in graded alcohols from 50% to 100% and in propylene oxide, and embedded in Epon. Embedded specimens were sectioned with a microtome and stained with toluidine blue for LM or with uranyl acetate and lead citrate for TEM. For scanning electron microscopy (SEM), after dehydration, the specimens were dried in a critical-point dryer, placed on aluminum stubs, and coated with gold.
RESULTS LIGHT MICROSCOPY
Despite the attempt to locate the intrastromal pockets at approximately half the depth of the cornea, light microscopic histologic findings showed that the hydrogel lenticules were actually situated from 3/4 to 4/5 of the stromal depth (Fig 1). The deeper the implant, the more the posterior cornea was seen to protrude into the anterior chamber. The protrusion was also seen on slitlamp examination in the monkey eyes, but was more pronounced in the histologic specimens. The epithelial cell layer in all the specimens was thinner in the central area over the implant, compared to the periphery outside the limits of the pocket and lenticule. In the central area, the epithelial cells were flat and irregular in shape. Bowman's membrane and the anterior stroma looked 1171
OPHTHALMOLOGY
•
OCTOBER 1984 •
VOLUME 91
•
NUMBER 10
Implant
AC Fig 2. Light micrograph. In the area of the protrusion beneath the edge of the hydrogel lens implant (implant), there is a fold in Descemet's membrane and some endothelial vacuolations can be seen (small arrows). Between the edges of the implant and the stroma, newly formed collagen fibrils are seen (large arrow). AC = anterior chamber (Toluidine blue, X287).
Fig 3. Epithelial cells in the central cornea are irregular in shape and have many vacuolations (V). BM = Bowman's membrane. (TEM, X7llO; bar = 2.8 JLm).
normal. No inflammatory cells were seen in the vicinity of the implant. Between the edges of the implant and the stroma, newly formed collagen fibrils and some keratocytes were seen. The posterior cornea appeared normal. Vacuolations were seen more frequently in some endothelial cells in the central area, compared to the periphery. Beneath the edges of the implant, where the posterior corneal protrusion into the anterior chamber began, a fold in Descemet's membrane and some vacuolations in the endothelial cells were seen (Fig 2). ELECTRON MICROSCOPY
The epithelial cells in the central cornea were irregular in shape and contained many vacuolations (Fig 3). The epithelial cells in the periphery looked normal, except for some edematous rough endoplasmic recticulum (Fig 4). In the anterior stroma over the implant, the collagen lamellae were slightly irregular and the cytoplasm in some keratocytes appeared to be of non-uniform density, indicating the early stages of degeneration (Fig. 5). At the outer edges of the intrastromal pocket, where the implant did not entirely fill the space made by the spatula at the time of surgery, fibrinous tissue in an amorphous matrix and some degenerated keratocytes were seen to form a connecting layer about 105 to 180 nm thick between the anterior and the posterior cornea (Fig 6). A similar material was seen between the implant and some stromal keratocytes immediately adjacent to the implant both anteriorly and posteriorly (Fig 7). Relatively more amorphous matrix was found between implant and keratocyte; more fibrinous tissue (55-140 A) was found between keratocyte and stroma. These keratocytes were sparsely distributed both anteriorly and posteriorly in the central area, but increased in numbers toward the periphery, where some accumulation of these lin
Fig 4. In the periphery, the epithelial basal cells look normal except for edematous rough endoplasmic reticulum (small arrows). The large arrow indicates a ferritin deposit in the intracellular space. Note the same material in the cy1oplasm. BM = Bowman's membrane; N = nucleus. (TEM, X7llO; bar = 2.8 JLm).
cells was seen. Macrophage and fibroblastic cells were observed between the implant and the stroma, also both anteriorly and posteriorly (Fig 8). The posterior cornea beneath the implant looked normal. Centrally, the endothelial cells were flat and showed some cytoplasmic vacuolations (Fig 9). Newly formed abnormal collagen fibrils were seen in the posterior Descemet's membrane. Scanning electron microscopy of the endothelial cells in the central area showed some edematous cells, as well as some with depressions in the middle and at the cell borders (Fig 10). The peripheral endothelial cells looked more normal by both transmission (Fig II) and scanning (Fig 12) electron microscopy, except for slight depression of the cell borders (Fig 12). No abnormal collagen fibrils were seen in the posterior Descemet's membrane peripherally.
YAMAGUCHI , et al
• HYDROGEL IMPLANTS IN PRIMATES
Fig S. Keratocyte (K) in anterior cornea over the implant. Nonuniform cytoplasmic density indicates early stages of degeneration. (TEM, X 10,428; bar = 1.9 I'm).
Fig 7. Interface between the posterior cornea (Post. St) and the hydrogel lens (implant). Newly formed fibrinous tissue in basement membranelike amorphous matrix (arrows) can be seen between the hydrogel lens and a keratocyte (K) and between a keratocyte and the posterior cornea; * indicates separation from posterior cornea during dissection (TEM, X47400, bar = 0.4 I'm).
DISCUSSION McCarey and Andrews l2 implanted 71 % water content lenses in rabbit corneas and found slowly progressive thinning of the stroma anterior to the edges of the impla.nt. They suggested that this complication was a result of mechanical stress. However, in our study, more abnormalities appeared in the central areas above and below the implant, and the character of these abnormalities indicated some type of physiological stress. The epithelial layer was thinner and the cells were more irregular in shape and had more vacuo lations over the implant, compared to the periphery (Figs 3, 4). Hamano et al. 10 also reported abnormalities of the corneal epithelium over intrastromal hydrogel implants observed by scanning electron microscopy. Also, we observed
Fig 6. Junction between anterior cornea (Ant. St) and posterior cornea (Post. St) at edge of intrastromal pocket. The wound, which was made by the spatula, is filled with amorphous matrix (arrows). A degenerated keratocyte (K) is also seen. (TEM, X28440; bar = 0.7 I'm).
Fig 8. Area between the hydrogel lens implant (implant), anterior cornea (Ant. St), and posterior cornea (Post. St); * indicates artifactitious separation during dissection. Note macrophage (large arrow) and fibroblasts (small arrow). (TEM, X711O; bar = 2.8 I'm).
abnormal collagen fibrils, presumably secreted by the vacuolated endothelial cells, on Descemet's membrane (Fig 9). Furthermore, some keratocytes immediately anterior to the implant displayed degenerative changes (Fig 5). In contrast, the peripheral areas of the cornea were more normal looking. It may be that the implant produces some endothelial changes, as well as impeding the nutrient flow from the aqueous humor toward the epithelium, resulting in the observed abnormalities. Although the posterior protrusion into the anterior chamber appeared to be more pronounced with more posterior placement of these implants, a more anterior placement would seem likely to carry an increased risk of anterior perforation or extrusion. Inasmuch as the procedure used here did not include a microkeratome lamellar dissection, no refractive increase was obtained, because of the barrier nature of Bowman's membrane and the anterior collagen lamellae, which prevents an1173
OPHTHALMOLOGY
•
OCTOBER 1984
•
VOLUME 91
•
NUMBER 10
AC Fig 9. Vacuoles (V) can be seen in the endothelial cells in the central cornea beneath the implant. Note newly formed abnormal collagen fibrils (arrows) on posterior Oescemet's membrane (OM). N = nucleus; m = mitochondria; AC = anterior chamber. (TEM, X16590; bar =
Fig 10. In the central cornea beneath the implant, the middles and cell borders of the endothelial cells are depressed (arrows). (SEM, X2152; bar = 4.6 I'm).
1.2 I'm).
AC Fig 11. In the periphery beyond the edges of the implant, the endothelial cells are more normal looking than the cells in the central area (Fig 9). The cristae of the mitochondria (m) are not clear. No abnormal collagen fibrils can be seen in the posterior Descement's membrane. OM = Descemet's membrane; AC = anterior chamber; N = nucleus. (TEM, X16590; bar = 1.2 I'm).
terior bowing of the cornea. However, the question of placement depth may apply to the keratophakia procedure as well, and a 50% depth has been suggested to be optimal by Werblin et al. 16 Our experience in this study was that although the pockets were created at 50% depth the implants were actually located from 3/4 to 4/5 the depth of the cornea. This problem clearly requires further work to ascertain the importance of implant depth and greater reliability of placement. In our study, keratocytes were seen in the stroma adjacent to the implant, distributed somewhat thinly centrally and more densely in the periphery. Also, keratocytes and newly formed fibrinous tissue in a basement-membrane-like amorphous matrix were ob1174
Fig 12. Endothelial cells look normal in the periphery except for the depressions at the cell boundaries (arrows). (SEM, X2182; bar = 4.6 I'm).
served between the implant and the stroma. Sendele et al 17 reported a case of hydrogel implantation in a human eye. Collagenous filaments and lined fibroblasts were seen at the implant/stroma interface six months after implantation, but no basement-membrane-like material was observed in the space between the plastic and the cornea. However, there appeared to be an artifactitious separation between the implant and the tissue, so such material may have been lost in the processing of the specimens. In any case, it appears that in both primate and human eyes, keratocytes play some role in secreting material to fill the spaces between the tissue and the plastic. No inflammatory reaction was seen around the implant, either in primate eyes in our study, or in the human eye examined by Sendele et al. 17 Plastic implants have considerable potential for mitigating some of the problems connected with current refractive surgical pro-
YAMAGUCHI, et al
•
HYDROGEL IMPLANTS IN PRIMATES
cedures, but further investigation of corneal tolerance for these synthetic materials will be needed to ensure long-term safety.
ACKNOWLEDGMENT The authors thank Mr. Hewitt Cabirac, of the Delta Regional Primate Research Center, Covington, Louisiana, for his assistance with the monkeys used in this study.
REFERENCES 1. Bock RH, Maumenee AE. Corneal fluid metabolism; experiments and observations. Arch Ophthalmol 1953; 50:282-5. 2. Stone W Jr, Herbert E. Experimental study of plastic material as replacement for the cornea; a preliminary report. Am J Ophthalmol 1953; 36(P!. 2): 168-73. 3. Krwawicz T. New plastic operation for correcting the refractive error of aphakic eyes by changing the corneal curvature; preliminary report. Br J Ophthalmol 1961; 45:59-63. 4. Knowles WF. Effect of intralamellar plastic membranes on corneal physiology. Am J Ophthalmol 1961; 51 :1146-56. 5. Belau PG, Dyer JA, Ogle KN, Henderson Jw. Correction of ametropia with intracorneal lenses; an experimental study. Arch Ophthalmol 1964; 72:541-7.
6. Onuma M, Hayashi T. Intralamellar transplantation of acrylic meshes and discs. Jpn J Clin Ophthalmol 1963; 17:453-7. 7. Brown SI, Dohlman CH. A buried corneal implant serving as a barrier to fluid. Arch Ophthalmol 1965; 73:635-9. 8. Dohlman CH, Refojo MF, Rose J. Synthetic polymers in corneal surgery. I. Glyceryl methacrylate. Arch Ophthalmol 167; 77:252-7. 9. Otsuka J, Tokoro T, Tarumi N, et al. On the influences hydrogel membranes on the cornea of rabbits. J Jpn Contact Lens Soc 1971; 13:140-4. 10. Hamano H, Hori M, Hirayama K. Scanning electron microscope observation of corneal epithelium. Part VI. Intralamellar implant of hydrogel contact lens. J Jpn Contact Lens Soc 1973; 15:111-22. 11. Mester U, Heimig 0, Dardenne MU. Measurement and calculation of refraction in experimental keratophakia with hydrophilic lenses. Ophthalmic Res 1976; 8:111-6. 12. McCarey BE, Andrews OM. Refractive keratoplasty with intrastromal hydrogel lenticular implants. Invest Ophthalmol Vis Sci 1981; 21 :10715. 13. Binder PS, Deg JK, Zavala EY, Grossman KR. Hydrogel keratophakia in non-human primates. Curr Eye Res 1982; 1:535-42. 14. Sher JH, Dixon WS. Interlamellar refractive keratoplasty in rabbits. Can J Ophthalmol1982; 17:116-20. 15. McDonald MB, Koenig SB, Friedlander MH, et al. Alloplastic epikeratophakia for the correction of aphakia. Ophthalmic Surg 1983; 14:65-9. 16. Werblin TP, Blaydes JE, Fryczkowski AW, Peiffer R. Stability of hydrogel intracorneal implants in non-human primates. CLAO J 1983; 9:157-61. 17. Sendele DO, Abelson MB, Kenyon KR, Hanninen LA. Intracorneal lens implantation. Arch Ophthalmol 1983; 101 :940-4.
1175