Characteristics of Corneal Xenograft Rejection in a Discordant Species Combination

Joel R. Ross, David N. Howell, and Fred P. Sanfilippo

Purpose. To characterize the fate of Lewis rat corneas transplanted to Hartley guinea pigs. Methods. Full-thickness Lewis rat corneal buttons were grafted orthotopically to Hartley guinea pigs (xenografts), ACI rats (allografts), or Lewis rats (isografts). Two panels of recipients were presensitized with xenogeneic skin grafts or allogeneic skin grafts. Serum samples were collected pre- and post-transplant and analyzed by flow cytometry and indirect immunofluorescence. Results. Unlike vascularized xenografts that reject within 30 min, corneal xenografts had a mean survival time of 8 days. Presensitization with guinea pig skin grafts increased recipient IgM and IgG xenoantibody levels, as measured by flow cytometry on guinea pig hematopoietic cells, and significantly accelerated corneal xenograft rejection with a mean survival time of 5 days. Presensitization with allogeneic ACI skin grafts had no effect on xenoantibody levels or xenogeneic corneal graft survival. Guinea pig corneas stained by indirect immunofluorescence with normal rat serum exhibited low (1+) but significant binding of IgG and IgM, primarily on epithelium and stroma. Serum from Lewis rats that rejected a corneal xenograft had elevated IgG and IgM xenoantibodies that reacted strongly (4+) with guinea pig cornea and heart. Conclusions. In the discordant guinea pig-to-rat species combination, donor corneas express xenoantigens; rejection of corneal xenografts stimulates IgM and IgG xenoantibody production; sensitization to xenoantigens can accelerate corneal xenograft rejection; and discordant corneal xenografts, unlike vascularized organs, are not hyperacutely rejected. Invest Ophthal- mol Vis Sci. 1993:34:2469-2476.

1 he shortage of human organs has led to a renewed ents (limited in number) are considered unlikely interest in the possibility of using xenogeneic tissues candidates as donors of xenogeneic tissues for a vari- for transplantation. Potential xenogeneic donors can ety of practical and ethical reasons.3"6 The possibility be subdivided into "discordant" donors, against of using discordant tissue sources for xenotransplan- whose tissues the recipient possesses preformed xe- tation is, then, of considerable importance. noreactive natural antibodies (NAb), and "concor- The major obstacle to transplantation of vascular- dant" donors, against whose tissues such NAb are ab- ized xenogeneic organs in discordant species combina- sent.12 Primate species concordant with human recipi- tions is hyperacute rejection. This rapid form of rejection, which typically occurs within minutes to a few hours of transplantation, is generally thought to be From the Departments of Pathology, Immunology, ami Surgery, Duke University and VA Medical Centers, Durham, North Carolina. mediated by NAb via activation of the classical com- Submitted for publication July 22, 1992; accepted December 21, 1992. plement (C) pathway, though direct activation of the Proprietary interest categoiy: N. Reprint requests: Fred Sanfilippo, M.D., Ph.D., Johns Hopkins Medical alternative C pathway appears to play a role in some Institutions, Department of Pathology, 600 N. Wolfe Street, Baltimore, MD 21205. species combinations.12'78 Because orthotopic cor-

Downloaded from iovs.arvojournals.org on 09/25/2021 2470 Investigative Ophthalmology & Visual Science, July 1993, Vol. 34, No. 8

neal transplants are placed in an avascular graft site, nized by corneal xenografts has elevated levels of IgM the role of these humoral rejection mechanisms may and IgG xenoantibody. be considerably diminished. Therefore, the transplantation of xenogeneic corneas may prove to be more successful than that of vascularized organs. METHODS Worldwide, the shortage of donor corneas has Animals been estimated to be as high as 10 million. The lack of corneas stems from many factors, including religious All animals involved in this study were cared for and or ethnic beliefs, lack of adequate eye banking facili- used in accordance with the ARVO Resolution on Use 1 ties, and poor quality of locally donated corneas. In of Animals in Research. Male Lewis (RT1 ) rats and the U.S., the need for corneas is not as acute, but Hartley guinea pigs were obtained from Charles availability varies from region to region. Potentially, as Rivers Laboratories (Wilmington, MA). ACI (RTla) with other tissues and organs, the use of xenogeneic rats were bred at the VA Medical Center Animal Re- donors may be a way to alleviate the shortage of donor source Facility. corneas. Many early studies of corneal xenografts exam- Orthotopic Corneal Transplantation ined the survival of heterotopic9 or intralamellar Full-thickness, 3.5-mm penetrating corneal grafts grafts,10"12 which have only limited clinical use and are were performed as previously described.23 Briefly, ani- not subjected to the same exposure to the host im- mals were anesthetized by intramuscular injection of mune system as are full-thickness grafts. Only a few with a combination of 100 mg/kg of a 100 mg/ml so- investigators have used a penetrating keratoplasty lution of ketamine (Parke-Davis, Morris Plains, NJ) technique to perform xenografts (Table 1). 13-21 Of and 2 mg/kg of a 100 mg/ml solution of xylazine (Mo- these, none have used animal species associated with bay, Shawnee, KS). Topical anesthesia, proparacaine well-established models of xenotransplantation. Con- (Alcon, Ft. Worth, TX) was applied to the eye. A sub- cordant models of corneal xenotransplantation have conjunctival injection of 0.05 ml of a solution contain- yielded promising results18"21 but, as mentioned, are ing 1 mg/ml atropine (Elkins-Sinn, Cherry Hill, NJ) of little practical usefulness for human corneal trans- and 1:1000 epinephrine (Parke-Davis) was used to di- plantation. late the iris. A 3.5-mm trephine was used to score the We have examined the survival of guinea pig cor- donor cornea, and the donor graft was removed using neas transplanted to Lewis rat recipients. This discor- curved corneal scissors. The recipient cornea was simi- dant species combination has been widely used to ex- larly scored with a 3.0-mm trephine, and the central amine the hyperacute rejection of vascularized xeno- 3-mm area was removed. The donor graft was then geneic transplants.1'28-22 Our findings indicate that, in sewn in place using 12 interrupted sutures of 11-0 contrast to cardiac xenografts, which fail in 17 ± 4 nylon (Alcon). The anterior chambers of recipients min,22 guinea pig corneal xenografts are not hyper- were re-formed with sodium hyaluronate (Healon, acutely rejected in Lewis rats, and that rejection is sig- Pharmacia, Piscataway, NJ). Six sutures were removed nificantly accelerated when the recipients are presen- on postgraft day 5 and the remaining six on postgraft sitized with xenogenic (but not allogeneic) skin grafts. day 8. Topical antibiotic, tobramycin (Alcon), was ap- Antibody labeling studies demonstrate that normal rat plied immediately after surgery and on postoperative serum contains NAb that binds to antigens expressed days 4 and 7. No immunosuppressant agents were ad- on guinea pig cornea and that serum from rats immu- ministered at any time during the study.

TABLE l. Historical Development of Orthotopic Corneal Xenotransplantation Author Year Model Outcome Reference

Kissam 1844 Pig to human Failed 13 D'Amico 1969 Cat to rabbit 33% successful 14 Haq 1972 Fish to human Failed 15 Durrani 1974 Bovine to human 21% improved vision 16 Bahn 1982 Bovine to cat Failed 17 Moore 1987 Cat to rabbit Failed 18 Insler 1991 Human to primate 67% successful 19 Li 1992 Human to primate 55% successful 20 She 1992 Mouse to rat Successful 21

Downloaded from iovs.arvojournals.org on 09/25/2021 Rejection of Corneal Xenografts 2471

Grafts were observed every other day for the first ACI rats (allografts), or Hartley guinea pigs (xeno- week and at least twice a week thereafter. Grafts were grafts). Two grafts from each donor category were graded for clarity, edema, and vascularization. When performed on the same day. Graft survival was ana- at least two of the three criteria were moderate to se- lyzed by a Mantel-Haenzel program, and differences vere, the grafts were considered to be rejected. between the groups were analyzed by Student's Mest. Skin Grafts RESULTS Recipients were grafted orthotopically with 1 cm2 do- nor skin on the thorax, as described.24 Grafts were Corneal Grafts in Naive Recipients inspected visually, and the percent of visible epithelial The results of corneal grafts into naive recipients are necrosis was recorded. shown in Figure 1. All 8 isografts survived throughout Measurement of Xenoreactive Antibody the 30-day observation period. Five of 6 ACI allografts were rejected in this high-responder strain combina- Guinea pig cervical lymph nodes were removed and tion, with a mean survival time of 11 days. All 10 Hart- teased apart in phosphate-buffered saline containing ley guinea pig xenografts were rejected between 6 and 0.5% bovine serum albumin (Sigma, St. Louis, MO) 9 days, with an average survival time of 8 days. Though and 0.02% NaN3 (PBA). Cells were then washed three the survival of xenografts was significantly shorter times in PBA and resuspended at a concentration of 3 6 than that of both isografts (P < .001) and allografts (P X 10 cells/ml. Cells were stained with serial dilutions = .016), it was strikingly longer than the brief survival of serum using PBA as the diluent for 30 min at 4°C. time of heart grafts in the same species combination, After washing twice with PBA, cells were stained with a which survived 17 ± 4 min.22 The histologic features of 1:100 dilution of goat anti-rat IgM-phycoerythrin the rejected corneal grafts included a massive diffuse conjugate and goat anti-rat IgG-fluorescein conjugate infiltration of mononuclear cells and polymorphonu- (Jackson Immunoresearch, West Grove, PA) for 30 clear cells, accompanied by severe edema throughout min at 4°C. The cells were washed twice, fixed in 1% the stroma, destruction of the endothelial layer, and neutral buffered formalin, and analyzed on a FAC- epithelial sloughing (data not shown). An extensive Scan (Becton-Dickinson, Mountain View, CA). The analysis of the histopathologic characteristics of re- use of paired secondary antibodies allowed simulta- jecting corneal xenografts and the phenotypic compo- neous assessment of rat IgG (FITC channel) and IgM sition of the infiltrating cells is in progress. (PE channel). Data are reported as mode fluorescence channel. Corneal Grafts in Allosensitized Recipients Immunofluorescence Histology To determine if presensitization to allogeneic antigens would lead to cross-reactivity with xenogeneic anti- Guinea pig eyes or heart were embedded in gelatin gens expressed on the cornea or heightened immune and snap-frozen in a liquid nitrogen-immersed isopen- tane bath. Tissue blocks were stored at -70°C until 5- to 7-jum sections were prepared and mounted on gela- CORNEAL GRAFT SURVIVAL IN NAIVE RECIPIENTS tin-coated glass microscope slides. After acetone fix- 100 I I ation, sections were stained with dilutions of heat- inactivated rat serum, washed and incubated with allograft n = 6 FITC conjugated goat anti-rat IgM (Accurate, West- 80- xenograft n = 10 bury, NY) or sheep anti-rat IgG (Serotec, Oxford, En- isograft n = 8 gland). Sections were then examined using a Zeiss 60- D-7982 Oberkochen microscope (Carl Zeiss Inc., Thornwood, NY). ,,40 Experimental Design Lewis rats were divided into three groups of corneal 20- graft recipients. One group had no manipulation before corneal transplantation (naive recipients). The other two groups received orthotopic skin grafts from 5 7 9 11 13 Hartley guinea pigs (xenosensitized recipients) or day post-transplantation from ACI rats (allosensitized recipients). Ten to 12 FIGURE 1. Survival of corneal grafts in naive recipients. days after skin grafting, recipient rats received an Lewis rats received corneal grafts from Lewis rats (isografts), orthotopic corneal graft from Lewis rats (isografts), ACI rats (allografts), or Hartley guinea pigs (xenografts).

Downloaded from iovs.arvojournals.org on 09/25/2021 2472 Investigative Ophthalmology 8c Visual Science, July 1993, Vol. 34, No. 8

responsiveness in recipients, a group of Lewis recipi- CORNEAL GRAFT SURVIVAL IN XENOIMMUNIZED ents was preimmunized with fully allogeneic ACI skin RECIPIENTS grafts before receiving corneal transplants. As shown in Figure 2, allosensitization had no effect on the survival of Lewis isografts (P ) .05). Xenogeneic corneal grafts were also unaffected by allosensitization (P ) allograft n = 7 .05). However, rats that received ACI skin grafts exhib- xenograft n = 8 ited a trend of decrease in the mean survival time of isograft n = 5 subsequent ACI corneal allografts from 11 to 8 days (P = .06).

Corneal Grafts in Xenosensitized Recipients To determine if the prolonged survival of guinea pig corneal grafts was simply the result of the isolation of the avascular graft site from the immune system, a days post-transplantation group of Lewis recipients was preimmunized with FIGURE 3. Survival of corneal grafts in xenosensitized recipi- Hartley guinea pig skin grafts. These recipients had ents. Lewis rats were immunized with Hartley guinea pig high levels of circulating guinea pig reactive IgM and skin grafts 10 to 12 days before corneal transplantation. IgG as measured by fluorescence cytometry at the time they received a corneal graft (see below). Both corneal isograft and allograft survival were unaffected by xe- Xenoantibody Production nosensitization (P > .05) compared to naive recipients Serum samples from corneal graft recipients were heat (Fig. 3). However, xenogeneic corneal grafts were re- inactivated and used to stain guinea pig lymph node jected significantly more quickly (P < .001) in the xe- cell targets as a standard assay used to measure NAb.22 nosensitized recipients (5 days) than in the naive recipi- At time 0 before grafting, recipients had only low lev- ents (8 days), with signs of rejection occurring as early els of IgM and undetectable IgG NAb (Fig. 6). How- as the third postoperative day including mild-to-mod- ever, after rejection of the xenografts but not the allo- erate graft edema and opacification with significant grafts, the levels of both guinea pig reactive IgM and involvement of the recipient cornea as well (Fig. 4a). IgG antibodies were much higher (Fig. 6). Similar re- By contrast, allografts in xenosensitized recipients at sults were obtained when guinea pig platelets or eryth- day 3 were clear, with normal levels of trauma-asso- rocytes were used as targets (data not shown). Figure 7 ciated graft opacification at the wound margins and demonstrates that in the recipients presensitized by with the recipient iris and pupil easily seen through the guinea pig skin grafts, high levels of IgM and IgG were graft (Fig. 4b). A summary of survival in all 9 donor-represent in the circulation at the time of corneal trans- cipient combinations is shown in Figure 5. plantation.

CORNEAL GRAFT SURVIVAL IN ALLOIMMUNIZED Expression of Xenoantigens on Cornea RECIPIENTS To determine if the prolonged survival in naive recipients of xenogeneic corneal grafts, relative to cardiac allograft n = 5 grafts, was the result of a paucity of xenoantigens on xenograft n = 10 the donor cornea, sections of normal guinea pig cor- isograft n = 5 nea were stained with serum from corneal graft recipients. Lewis rats that had rejected xenogeneic corneal grafts contained serum IgG and IgM, which reacted strongly (4+) with guinea pig cornea (Figs. 8a and 8b). Epithelial cells exhibited moderate diffuse staining with a superimposed brighter, granular staining pattern. Stromal staining was also present and was con- fined to cellular elements. No staining of endothelium days post-transplantation or of the endothelial or epithelial basement mem- FIGURE 2. Survival of corneal grafts in allosensitized recipi- branes was detected. Serum from Lewis rats that re- ents. Lewis rats were immunized with ACI rat skin grafts 10 jected allogeneic corneal grafts displayed weak but sig- to 12 days before corneal transplantation. nificant staining (1+), with a distribution similar to

Downloaded from iovs.arvojournals.org on 09/25/2021 Rejection of Corneal Xenografts 2473

A FIGURE 4. Clinical appearance of corneal grafts in xenosensitized recipients, (a) On the third postoperative clay, a xenograft is already completely opaque with moderate-to-severe edema, (b) An allograft at the same time is still completely clear with an easily detectable pupil.

that seen for xenogeneic corneal recipients. However, with serum from allogeneic graft recipients or naive an identical pattern and intensity of staining was ob- rats. tained when serum from naive rats was used (Fig. 8c), indicating that reactivity was against antigens recog- DISCUSSION nized by natural antibodies. When sections of guinea pig heart were stained Vascularized discordant xenogeneic transplants are with serum from recipients of allogeneic and xeno- rejected hyperacutely within minutes to hours. By geneic corneal grafts, a vascular staining pattern was contrast, we found that orthotopic corneal grafts in observed (Fig. 9). Again, sections stained with the the well-characterized guinea pig-to-Lewis rat species serum of a xenograft recipient were more intense combination survived for up to 8 days before being than, but showed the same pattern as, sections stained

XENOREACTIVITY OF SERUM FROM NAIVE CORNEAL GRAFT RECIPIENTS SURVIVAL OF CORNEAL TRANSPLANTS igM

ISOGRAFTS xeno 2 naive . n = 8 alloimmune . n = 5 allo 1 xenoimmune n = 5 liffllft ALLOG RAFTS allo 2 naive . n = 6 o o o o o alloimmune . n = 5 (X> O o xenoimmune n = 7 O SB XENOGRAFTS iso 2 naive n = 10 o 30 FIGURE 6. Xenoreactivity of serum from naive corneal graft SURVIVAL TIME (DAYS) recipients. Guinea pig lymph node cells were stained with FIGURE 5. Summary of guinea pig xenograft survival in serial dilutions of serum from corneal graft recipients and naive, xenosenisitized, and allosensitized Lewis rats. analyzed by flow cytometry.

Downloaded from iovs.arvojournals.org on 09/25/2021 2474 Investigative Ophthalmology & Visual Science, July 1993, Vol. 34, No. 8

XENOREACTIVITY OF SERUM not express the same or as many xenogeneic antigens. FROM XENOIMMUNIZED RECIPIENTS However, inflammation after corneal transplantation can lead to upregulation of class II major histocompat- IgM ibility complex antigens on corneal tissues,25 and per- —•- xeno 1 eoo- "gG / 600- i / haps the same is true of xenoantigens. Thus, a grafted (D «*• guinea pig corneal endothelium may express more tar- ^- i CD —B— xeno 2 f 600- gets for NAb binding than a normal one. CO t (D 400- 1 I 500- / // Corneal grafts are exposed to immunoglobulins O -*- allo 1 3 / 400- through both the aqueous humor and tears; the po- ^300- d) // tential impact of this exposure may be lessened by lim- / / 300- aKJ o - / A/ -*- allo 2 ited quantities of antibody compared with vascularized £ 200-

100- grafts and limited exposure to cellular or humoral co- 100- -^ isol /^\^ factors such as C components. The importance of al- 0* 01 10 -8 -6 -4 -2 0 2 4 6 • 1^—-0 -a —-6 -4 -2 0 2 4 t ternative C pathway activation in this discordant t t t t model relative to direct activation by NAb8'22 may also skin graft carneal graft skin graft eorneal graft account for the prolonged survival of guinea pig cor- FIGURE 7. Xenoreactivity of serum from xenoimmunized re- neal xenografts by rats. cipients. Lewis rats received guinea pig skin grafts 10 days The prolonged survival of xenogeneic corneal before corneal transplantation. Guinea pig lymph node cells grafts holds promise for their future clinical applica- were analyzed as in Figure 6.

rejected by naive recipients even in the absence of any immunosuppression. Immunization of the recipients with allogeneic skin grafts did not affect the survival of subsequent xenogeneic corneal grafts. However, preimmunization with xenogeneic skin grafts stimu- lated xenoantibody responses and significantly short- ened survival times of xenogeneic corneal grafts from 8 days to 5 days. This indicates that the prolonged survival of xenogeneic corneal grafts is not merely the result of isolation from the recipient immune system and that xenografts can be rejected in an accelerated manner, faster than that previously reported for allografts in any experimental model. The lack of hyperacute rejection may be attrib- uted to the paucity of cornea-reactive NAb, or target antigens, to sequestration of the graft from NAb exposure, or to the absence of other proteins or cells that act as cofactors with NAb. Immunofluorescence studies demonstrated that the serum of naive rats and rats that received an allograft contained both IgM and IgG, which bound to guinea pig cornea, indicating that xenogeneic antigens recognized by NAb are expressed on the guinea pig cornea. The low level of staining suggests that the xenoantigens or NAb are present at low levels. By contrast, serum from rats that FIGURE 8, Expression of xenoantigens on the cornea, (a) Sec- rejected guinea pig corneal grafts stained normal tions of normal guinea pig cornea were incubated with guinea pig cornea much more strongly. The brightest serum from the recipient of a xenograft and stained with goat anti-rat IgG. Cellular elements in the epithelium and staining was observed on epithelial and stromal cells stroma stained positively (original magnification, 250x). (b) and represented genuine antibody binding rather than A higher magnification of the epithelium is shown in this Fc receptor-mediated binding because the level of panel (original magnification, 400x). (c) Epithelium of a cor- binding was much higher than that obtained with naive nea incubated with serum from the recipient of an allograft or alloimmune sera. No endothelial staining was pres- and stained with goat anti-rat IgG (original magnification, ent, which suggests this critical layer of the cornea may 400x). Only faint staining is detectable.

Downloaded from iovs.arvojournals.org on 09/25/2021 Rejection of Corneal Xenografts 2475

FIGURE 9. Expression of xenoantigens on the heart. Sections of normal guinea pig heart were stained as in Figure 7, with (a) serum from a rat that had rejected a corneal xenograft (original magnification, 400x) or (b) an allograft (original magnification, lOOx).

tion. The 8-day survival time of these grafts is not in- ies of the feasibility of discordant cornea xenotrans- consistent with cellular rejection, for which well-estab- plantation. lished therapies already exist. The finding that presensitization to allogeneic antigens did not lead to Key Words accelerated rejection of subsequent xenografts is also corneal transplantation, xenotransplantation, hyperacute encouraging. Allosensitization, especially from prior rejection, antibody response graft rejection, has been reported to be associated with higher risk for graft rejection of vascularized or- Acknowledgment gans26 and corneal grafts.27'28 Xenografts, therefore, may represent a viable option for those patients who The authors thank Dr. William Baldwin III for technical ad- have already rejected a corneal allograft or who have vice and helpful discussions. been sensitized by pregnancy or transfusion. References In summary, we have examined corneal graft rejection in a well-characterized discordant rodent spe- 1. Auchincloss H. Xenogeneic transplantation. Trans- cies combination where hyperacute rejection of vascu- plantation. 1988;46:l-20. larized organs occurs. Our findings indicate that the 2. Bach F. Discordant xenografting: A summary and cornea expresses xenoantigens, that rejection of cor- look to the future. Transplant Proc. 1992; 24:739-742. neal xenografts promotes xenoantibody production, 3. Sanfilippo F. The use of non-human primates in xeno- and that presensitization to skin xenografts elevates transplantation. In: Hardy M, ed. Xenograft—25. New York: Elsevier; 1989:171-178. xenoantibody levels and significantly accelerates sub- 4. Caplan A. Is xenografting morally wrong? Transplant. sequent corneal xenograft rejection. Unlike vascular- Proc. 1992; 24:722-727. ized allografts, however, corneal xenografts are not 5. Najarian JS. Overview of in vivo xenotransplantation hyperacutely rejected, even in presensitized recipi- studies: Prospects for the future. Transplant Proc. ents. These findings provide the basis for further stud- 1992;24:733-738.

Downloaded from iovs.arvojournals.org on 09/25/2021 2476 Investigative Ophthalmology & Visual Science, July 1993, Vol. 34, No. 8

6. Singer P. Xenotransplantation and speciesism. Trans- 19. Insler MS, Lopez JG. Heterologous transplantation plant Proc. 1992;24:728-732. versus enhancement of human corneal endothelium. 7. Sachs D, Bach F. Immunology of xenograft rejection. Cornea. 1991; 10:136-148. Hum Immunol. 1990;28:245-251. 20. Li C, Xu J, Kong F, Li J. Experimental studies on pen- 8. Miyagawa S, Hirose H, Sirakura R, et al. The mecha- etrating heterokeratoplasty with human corneal grafts nism of discordant xenograft rejection. Transplanta- in monkey eyes. Cornea. 1992; 11:66-72. tion. 1988; 46:825-830. 21. She S-C, Moticka EJ, Feder J. Prolonged survival of 9. Duguid IGM, Koulmanda M, Mandel TE. Prolonga- xenogeneic corneal grafts following anterior chamber tion of heterotopic human corneal graft survival in injection of xenogeneic lymphocytes. Invest Ophthal- mice treated with an anti-CD4 monoclonal antibody. mol Vis Sci. 1992;33:832. Transplant Proc. 1990; 22:2107-2108. 22. Pruitt SK, Baldwin WM, Marsh HC, et al. The effect of 10. Versura P, Cellini M, Zucchini PG, Caramazza R, Las- soluble complement receptor type 1 on hyperacute chi R. Ultrastructural and immunohistochemical xenograft rejection. Transplantation. 1991;52:868- study of the effect of topical cyclosporin A in the rab- 873. bit eye. Cornea. 1989;8:81-89. 23. Ross J, He Y, Pidherney M, Mellon J, Niederkorn JY. 11. Basu PK, Ormsby HL. Corneal heterografts in rabbit. The differential effects of donor vs host Langerhans Am] Ophthalmol. 1957;44:477-480. cells in the rejection of MHC matched corneal allo- 12. Laino P, Mondino B. Immune response to organ-cul- grafts. Transplantation. 1991; 52:857-861. tured human cornea. Ann Ophthalmol. 1977; 9:83-87. 24. Billingham RE. Free skin grafting in mammals. In: Bil- 13. Kissam RS. Ceratoplastice in man. NY J Med. lingham RE, Silvers WK, eds. Transplantation of Tissues 1844;2:281-282. and Cells. Philadelphia: Wistar Institute; 1961:1-26. 14. D'Amico RA, Chaunico R, Castroviejo R. Suppression of the immune response to corneal xenografts. Trans- 25. Ross JR, Callanan D, Kunz H, Niederkorn JY. Evi- plant Proc. 1969; 1:256-258. dence that timing of class II expression determines 15. Haq M. Fish cornea for grafting. Br Med J. the fate of class II disparate corneal grafts. Transplan- 1972;5815:712-713. tation. 1991;51:532-536. 16. Bahn CF, MacCallum DK, Lillie JH, Meyer RF, Mar- 26. Sanfilippo F, Vaughn WK, Bollinger RR, Spees EK. tonyi CL. Complications associated with bovine cor- Comparative effects of pregnancy, transfusion, and neal endothelial cell-lined homografts in the cat. Invest prior graft rejection on sensitization and renal trans- Ophthalmol Vis Sci. 1982; 22:73-90. plant results. Transplantation. 1982;34:360-366. 17. Durrani KM, Kirmani TH, Hassan MM, et al. Pene- 27. Stark WJ, Taylor HR, Bias WB, Maumenee AE. Histo- trating keratoplasty with purified bovine collagen. compatibility (HLA) antigens and keratoplasty. Am J Ann Ophthalmol. 1974; 6:639-646. Ophthalmolmol. 1978; 86:595-604. 18. Moore MB, Gebhardt BM, Verity SM, McDonald MB. 28. Sanfilippo F, MacQueen JM, Vaughn WK, Foulks GN. Fate of lyophilized xenogeneic corneal lenticules in Reduced graft rejection with good HLA-A and B intrastromal implantation and epikeratophakia. Invest matching in high-risk corneal transplantation. N Engl Ophthalmol Vis Sci. 1987; 28:555-559. J Med. 1986;315:29-35.

Downloaded from iovs.arvojournals.org on 09/25/2021