Survival of Donor-Derived Cells in Human Corneal Transplants

Abstract

purpose. To determine the fate of donor epithelial, stromal, and endothelial cells after corneal transplantation in humans.

methods. Fifty-two transplanted corneal buttons were explanted over a 2-year period from patients who required regrafting and had received corneas from donors of opposite sex. Fluorescence in situ hybridization of the sex chromosomes of the epithelial, stromal, and endothelial cells was performed in histologic sections prepared from each freshly explanted graft. Fluorescence microscopy was subsequently used to determine the origin of cells in the graft (donor or recipient) and to quantify the relative proportion of donor and recipient cells of each corneal cell type.

results. As early as 3 months after transplantation, donor epithelial cells were completely replaced by recipient epithelium in all corneal buttons examined. Donor stromal and endothelial cells, however, were found in all 52 buttons, with 4% to 95% of stromal cells and 6% to 95% of endothelial cells being of donor origin. No significant correlation between donor cell proportion and the age of the graft could be found. Donor-derived cells were found in significant numbers up to 32 years after transplantation. Eight corneas in this study were transparent, compensated grafts, and a similar long-term survival of donor stromal and endothelial cells was found in these cases.

conclusions. Although donor epithelial cells are promptly replaced, a high proportion of donor stromal and endothelial cells can survive within the corneal transplant in the long-term. The proportion of surviving donor cells is highly variable; however, the source of this variability remains unknown.

In 1905, Zirm performed the first successful human corneal transplant. One hundred years later, corneal transplantation (penetrating keratoplasty) is a well-established and successful treatment modality for a range of corneal diseases. In the United States, it is estimated that approximately 50,000 surgeries are performed each year, whereas in Sweden, with a population of 9 million, 500 to 600 corneal transplants are performed annually, according to the Swedish Corneal Register. ¹ Although most corneal transplants are successful, an overall 2-year rejection rate of 15% in Sweden has been reported, ¹ and in so-called high-risk cases, the rejection rate can be much higher. ² These events underscore the need for a more complete understanding of the pathogenesis of immune rejection after keratoplasty in the normally avascular, transparent, immune-privileged cornea. Of particular interest in this regard are the interactions between the recipient cornea and the new graft at the cellular level where healing, antigen activity, and the ultimate transparency of the cornea are mediated. Unfortunately, these cellular interactions are not well understood. Even the basic question as to whether donor cells survive after transplantation is a fundamental biological problem that to this day remains unanswered. In fact, this important question has been the source of scientific discussion and debate ever since Zirm first showed that penetrating keratoplasty could be performed in humans. An excellent review of the literature relevant to this question was presented by Dohlman ³ and later by Wollensak and Green. ⁴ Depending on the technique used for analysis, some investigators have concluded that transplanted cells are replaced by cells from the recipient cornea, whereas others have claimed that donor cells in the transplanted tissue survive indefinitely. Without a sensitive and specific technique that could separate and unequivocally identify donor and host cells, both viewpoints could be argued. In a 1999 study by Wollensak and Green, ⁴ the technique of fluorescent in situ hybridization (FISH) analysis of the X and Y chromosomes was used for the first time to distinguish between individual recipient and donor cells in the human corneal epithelium, stroma, and endothelium with a high reliability in cases of sex mismatch between donor and recipient. In 14 failed, sex-mismatched grafts obtained retrospectively, Wollensak and Green found complete replacement of donor epithelium and endothelium by recipient cells in all grafts, whereas donor keratocytes were found in only three grafts, with a maximum survival time of 4.5 years. They concluded that “all cell types of corneal transplants tend to be replaced by recipient cells in the long term,” although “individual variability in the process of replacement exists.” Moreover, they proposed further studies investigating donor cell replacement in transparent, clinically successful transplants, as their study was limited to failed grafts. 

In the present study, we re-examined the question of donor cell survival in the corneal transplant by applying the FISH technique to a larger sample of freshly explanted corneal buttons at reoperation, a small subset of which were transparent, otherwise successful grafts removed for refractive reasons. 

Materials and Methods

Patients

With approval from the Gothenburg University ethics committee and according to the tenets of the Declaration of Helsinki, between February 2002 and January 2004, 60 corneal buttons were collected prospectively from patients with failed corneal transplants at the time of reoperation by members of the Swedish Society of Corneal Surgeons. All specimens were collected in sex-mismatched cases where the patient and the original donor were of different sex. Corneal button diameter varied from 7.25 to 8 mm. Immediately after surgery, the explanted buttons were fixed in formalin for 24 hours, then placed in 70% ethanol and sent to a single laboratory at Gothenburg University for histochemical preparation. Eight specimens were unsuitable for analysis, leaving a study sample of 52 corneal buttons. Recipients were distributed almost equally between women (n = 24) and men (n = 28), with a mean recipient age of 64.7 ± 15 years (mean ± SD) at the time of reoperation. The time from initial penetrating keratoplasty to removal of the original donor button at reoperation ranged from 3 months to 32 years. The most common indications for the primary transplant were keratoconus (22 cases) and corneal edema (12 cases). The main indication for reoperation was decompensation of the graft (described in clinical records as edema, endothelial rejection, or graft failure); however, 10 nonedematous, compensated grafts were removed for other reasons, eight of which were substantially or completely transparent. Details of the patients and the explanted corneal buttons are given in Table 1 . 

Histologic Pretreatment

In the laboratory, the formalin-fixed, ethanol-immersed corneal buttons were dehydrated and embedded in paraffin. Two full cross sections (5 μm thick) were cut from the center of each button and mounted on a glass slide. The sections were subsequently deparaffinized in xylene and rehydrated through a series of rinsing steps with decreasing concentrations of ethanol. Sections were then treated with 0.2 M HCl for 10 minutes at room temperature. After a rinse in PBS, the sections were treated with a proteinase K (Sigma-Aldrich, St. Louis, MO) solution (50 μg/mL) in 100 mM Tris-HCl with 50 mM EDTA (pH 8.0) at 37°C for 20 minutes. Enzyme digestion was stopped by immersing the slides for 5 minutes in 0.2% glycine in PBS at room temperature. After they were rinsed with distilled water, the sections were brought to 99% ethanol concentration through a series of rinsing steps in increasing concentrations of ethanol. FISH analysis of the sex chromosomes was then performed. The FISH technique has been described in detail elsewhere ⁵ and is outlined briefly. 

FISH Procedure

The samples were fixed in a 3:1 EtOH/acetic acid solution at room temperature for 15 minutes. FISH was performed with directly labeled DNA probes specific for the X and Y chromosomes (CEP-X/Y; Vysis, Inc., Abbott Laboratories, Downers Grove, IL). The DNA probe for chromosome X (DXZ1) was directly labeled with red fluorochrome (spectrum orange) specific for the AT rich α satellite DNA sequence at the centromeric region of chromosome X (Xp11.1-Xq11.1). The DNA probe for chromosome Y (DYZ1) was a collection of DNA segments (satellite III), directly labeled with green fluorochrome (spectrum green) and hybridized to most of Yq11.2 and all Yq12, the telomere of the Y chromosome. Probe hybridization was accomplished by placing 9 μL of the DNA probe mixture on each slide. An 18 × 18-mm coverslip was attached and sealed with rubber cement. Target and probe were denatured simultaneously at 80°C for 5 minutes on a heat block. Hybridization took place at 37°C in a moist chamber for 3 hours. The slides were washed at 42°C in 50% formamid/2× SSC (pH 7.6) for 15 minutes. If the coverslip had not fallen off after 2 minutes it was carefully removed. The slides were further washed in 2× SSC (pH 7.0) for 10 minutes and transferred to 0.01% Nonident (NP)-40 (pH 7.0) for 5 minutes. The samples were allowed to air dry and were counterstained with 9 μL of 4′,6-diamidino-2-phenylindole (DAPI II; Vysis Inc.). Coverslips were attached before analysis. 

Evaluation of FISH Signals

Only cells with two distinctive signals per nucleus were enumerated. Overlapping interphase nuclei or cells with an indistinct nuclear membrane were excluded from the evaluation. Signals of lower intensity were interpreted as minor binding sites (cross-hybridization) and similarly ignored. Paired fluorescent spots were counted as one signal when appearing less than one signal diameter from each other. Less than 2% cells with only one signal present is a realistic standard of acceptance for the DNA labeling procedure (Vysis, Inc.). The analysis was performed on a fluorescence microscope (Nikon, Tokyo, Japan) equipped with a digital camera for image capture. An X chromosome centromere exhibited an orange signal, a Y q-arm exhibited a green signal, and the nucleus was counterstained with DAPI-blue. A triple bandpass DAPI/FITC/TRITC filter (360/490/570 nm) was used to view all three fluorescent signals simultaneously. 

All cell counts were performed by a single observer, and two transverse sections from each corneal button were used for cell counting. Cell nuclei were identified as being male- or female-derived, based solely on the presence of two orange signals (female) or a green and orange signal (male) within a single nucleus. At least 100 nuclei from the epithelium and at least 100 nuclei from the stroma were counted from each button. The endothelium, however, was missing in eight specimens and fewer than 30 endothelial cells were found in nine specimens; so, these buttons were excluded from the endothelial analysis. In the remaining 35 corneal buttons, between 30 and 100 endothelial cell nuclei from each button were counted for the analysis. 

Cell counting results were tabulated in a spreadsheet, and the Student’s t-test was used to compare means. Linear relationships among variables were analyzed with the Pearson correlation coefficient. In all cases, P < 0.05 was considered significant. All statistics were performed with spreadsheet software (Excel 2003; Microsoft, Inc., Redmond, WA). 

Results

No donor-derived epithelial cells were detected in any of the corneal buttons, whereas all 52 corneas had some donor-derived stromal cells (keratocytes) and 26 of 35 corneas had some donor-derived endothelial cells (Table 1 , Fig. 1 ). Donor cells were distributed throughout the corneal sections in a seemingly random fashion. 

Donor-derived keratocytes were found with a proportion ranging from 4% to 95%, with no significant correlation of donor keratocyte survival with graft age (Fig. 2) . The lack of correlation with graft age (time period within the recipient cornea) persisted after stratification of the data along sex, graft transparency (8 cases), recipient age, and indication for the initial transplant. In 33% (17) of cases, donor keratocytes persisted for more than 10 years after, and in one patient with a transparent graft, 65% of keratocytes counted were donor-derived 32 years after surgery. 

Analysis of endothelial cells from 35 corneal buttons revealed 9 cases in which donor endothelial cells were completely replaced by recipient endothelium, 24 cases in which donor and recipient endothelial cells coexisted, and 2 cases in which only donor-derived endothelium was present (Table 1) . In cases in which both donor and recipient endothelial cells were found, the proportion of donor-derived cells ranged from 6% to 95%. Similar to donor keratocytes, donor endothelial cell survival was not significantly correlated with age of the graft (Fig. 3) , regardless of transparency or indication for the initial transplant. When stratified by sex and recipient age, however, a significant decline in the frequency of donor endothelial cells with increasing postoperative time was found in the women (P = 0.04 , 14 corneas) and in patients older than the mean age of 64.7 years at the time of explantation (P = 0.03, 16 corneas). In addition, these variables were correlated as female recipients in this study (70.6 ± 10 years, mean ± SD) were significantly older than the male recipients (59.6 ± 17 years; P = 0.006). 

Ten of the explanted corneal buttons were removed for reasons other than edema. Of these, two were removed due to suture-related infection, five were completely transparent buttons removed due to intractable astigmatism, and three were transparent except for small, localized scars on the visual axis. Eight grafts therefore exhibited general stromal transparency and endothelial compensation. As seen in Table 1and in Figures 2 and 3 , the frequency of donor-derived keratocytes and endothelial cells in these eight grafts did not generally differ from those of edematous, decompensated grafts. No clear trend in donor cell survival with graft age was noted in this small sample of transparent grafts. 

Keratoconus patients in this study were significantly younger (54.5 ± 14 years) than those with edema as an initial indication for transplantation (72.1 ± 13 years; P = 0.001), and the survival time of the original transplant was significantly longer with keratoconus (158.9 ± 104 months) than with edema (27.8 ± 16 months; P < 0.001). Similarly, the seven patients with pseudophakic bullous keratopathy as an initial indication for transplantation tended to be older (75.6 ± 8 years) with a relatively short time interval until reoperation (36.7 ± 24 months). Although the seven patients with endothelial dystrophy also tended to be older (69.1 ± 9 years), transplant survival time varied widely; however, in all seven of these cases the proportion of surviving donor keratocytes was high at the time of explantation (66%–93%). 

Discussion

We have successfully performed the largest study to date using the FISH technique to analyze human corneal donor buttons after penetrating keratoplasty. In all 52 corneal buttons analyzed, complete epithelial replacement by recipient cells was observed, as early as 3 months after surgery in one case and after 6 months in two cases. In earlier studies using the FISH technique, complete epithelial cell turnover was similarly observed to occur within 6 months to 1 year after transplantation. ^{4 6 7} Although donor-derived epithelial cells have been shown to persist for years after limbal transplantation in patients with limbal stem cell deficiency, ^{6 7 8 9} patients in this study did not have initial transplant indications consistent with limbal stem cell deficiency, and it is therefore not surprising that complete replacement of donor epithelial cells occurred in the promptly re-epithelialized graft. Confirmation of the recipient sex in all 52 sex-mismatched cases by epithelial cell analysis of the donor buttons served to validate the FISH procedure in this study in the absence of sex-matched control cases. 

Although donor epithelial cells were quickly and completely replaced, the hypothesis that donor keratocytes and endothelium are gradually replaced over time ^{3 4} could not be supported by the results of this study. Long-term survival of donor keratocytes and endothelium was observed despite a substantial variability in patient characteristics and presumed transplant conditions (initial operations having been performed across multiple centers over a 32-year period). Although a tendency for donor endothelial cells to be replaced over time was observed in patients older than 65 years, even within this group, the long-term survival of a substantial proportion of donor endothelial cells was found in some patients (Table 1) . Furthermore, the presence of eight transparent grafts in this study—exhibiting similarly variable long-term survival of donor keratocytes and endothelial cells—suggests that donor cells may survive in the long term in most corneal transplants, including successful ones. 

The results of this study differ most notably from those reported by Wollensak and Green, ⁴ who concluded that all cell types of the corneal transplant tended to be replaced by recipient cells in the long term. Both the small sample size that was used (14 buttons) and the retrospective nature of the study (where archived samples were used for analysis) may account in part for the discrepancy with the present results. In the present study, corneal buttons were obtained prospectively, with fresh specimens used for analysis in all cases. The time from removal of the original donor button to cellular analysis did not exceed more than a few days. 

Our findings support the contention that donor keratocytes and endothelial cells can survive in the transplanted cornea indefinitely, a theory supported by earlier animal experiments. ^{10 11 12 13 14 15} Before this study, the longest period donor keratocytes have been reported to survive within a transplanted cornea was 6.5 years ¹⁶ ; Our findings suggest that individual keratocytes (or keratocytes derived from a single source cell) may remain in the corneal stroma for life. Although damaged donor keratocytes can be replaced by cell division in primates, ¹⁷ it is unclear whether the donor keratocytes observed in this study were original or had been replenished through cell division. 

As endothelial cells do not generally proliferate mitotically, ¹⁸ the donor endothelial cells were probably original, and some of them may be expected to persist in the transplant for life. These findings contradict those of Espiritu et al., ¹⁹ who observed complete replacement of donor endothelial cells in rabbits after 7 months, using the sex chromatin method, and Dohlman ³ who reported the disappearance of isotopically labeled endothelial cells in rabbits 10 days after transplantation. Methodological limitations in these earlier studies have been noted. ^{10 11} The findings in the present study are in agreement with experiments conducted by several investigators, ^{10 11 14} who observed donor endothelial cell survival in rabbits for 12 to 21 months by using both radiolabeling and sex chromatin methods. 

Previous studies of corneal transplantation in cats, ²⁰ rabbits, ^{11 13 14 15 21} and more recently in mice, ²² have suggested that the long-term presence of donor-derived endothelial cells and keratocytes may be a necessary condition for graft transparency and long-term survival, as opaque grafts are typically characterized by invading recipient cells, vascularization, and a conspicuous paucity of donor cells. Our observations of significant proportions of surviving donor keratocytes and endothelial cells in grafts having remained transparent and avascular for long periods appear to support the observations of these earlier studies. Notably, in one study, Polack et al. ¹¹ used radiolabeling to study donor cell survival in grafted rabbit corneas and concluded that the scar tissue in failed, opaque grafts was primarily of host origin. In grafts that remain transparent and avascular, it has been speculated that the immune privilege of the cornea may protect donor cells for long periods within the graft, ²⁰ whereas in highly vascularized tissue, such as skin grafts, donor cells do not survive. ^{4 11}  

Another possible explanation for the long-term persistence of donor cells that cannot be excluded is the potential existence of donor stem cells within the graft. Although the presence of stem cells was not investigated in this study, the future testing of explanted buttons with suitable stem cell markers could address this hypothesis. 

Although in the present study donor keratocytes and endothelial cells survived in the long term, the proportion of surviving donor cells was subject to considerable individual variability. Although recipient age, sex, indication for transplantation, graft survival time, and transparency of the graft did not appear to affect the proportion of donor cells that survived, other variables such as donor characteristics (donor age, time in culture, corneal preservation method), the type and degree of surgical trauma, preoperative risk level, and postoperative complications, vascularization, and rejections may help to explain the wide variability in donor cell proportion observed, although detailed data were unfortunately not available in the present study. Future studies controlling for these variables may help to determine why a large proportion of donor cells thrive in some transplants, whereas their numbers are severely diminished in others. 

Our finding of significant populations of both donor and recipient keratocytes and endothelial cells within the graft in the long term is positive evidence of cell migration into the graft, but instead of complete donor cell replacement, we speculate that keratocytes and endothelial cells of both the donor and recipient may reach equilibrium within the transplanted cornea in the long term, although the relative proportion of surviving donor cells may be subject to considerable individual variability. Factors such as immunologic reaction, postoperative complications, and certain dystrophies present in the recipient cornea may upset this equilibrium and tip the balance in favor of the recipient cells. 

Several limitations to the present study are worth noting. The use of only a few histologic sections from the central region of each corneal button provided us with only a small sampling of the cell population within the transplanted cornea. That these sections are representative of the entire corneal button was implicitly assumed; nonuniformity, however, in the distribution of donor and recipient cells (with more cells of one type appearing at the graft edge, for example) would affect the relative proportion of donor cells detected. Future studies using either numerous sections taken across the entire corneal button or a flatmount technique would provide a more accurate indication of the distribution of surviving donor cells. Another limitation in this study was the presence of only a small number of nonedematous, transparent grafts. The detection of significant differences in donor cell survival between transparent (i.e., clinically successful) and failed transplants could yield insights into the pathogenesis of graft failure; however, such an analysis was not possible in this study as it would require a larger group of clinically successful transplants removed for refractive reasons. Finally, although a relatively large number of excised corneal buttons were examined in this study, the cross-sectional nature of the samples resulted in missing or incomplete medical information about the donor, surgical conditions, and postoperative period in many cases. Such information may be critical for determining the factors that correlate with or confound the observed degree of donor cell survival. 

In summary, the results of this study indicate that, although donor epithelial cells in corneal transplants are promptly replaced by recipient epithelium, both keratocytes and endothelial cells from the donor can survive in significant proportions within the transplanted human cornea in the long term (up to 32 years), with no detectable trend toward replacement by recipient cells over time. The proportion of donor-derived keratocytes and endothelial cells surviving within the corneal transplant is subject to considerable individual variability, however, no source for this variability could be found. Examination of eight transparent grafts in this study revealed a similar variability in donor cell survival. Further studies are needed to investigate the origins of the wide variability in donor cell survival after corneal transplantation and may help to elucidate the relationship between donor cell survival, immunologic activity, and the ultimate success of the corneal transplant. 

Appendix

Swedish Society of Corneal Surgeons Study Participants

Ingrid Floren, Department of Ophthalmology, Lund University Hospital, Lund, Sweden; Per Fagerholm, Ulf Wihlmark, Björn Lundh, Department of Ophthalmology, Linköping University Hospital, Linköping, Sweden; Ulf Stenevi, Margareta Claesson, Hussein Shams, Department of Ophthalmology, Sahlgren University Hospital, Mölndal, Sweden; Helena Sönne, Department of Ophthalmology, Örebro University Hospital, Örebro, Sweden; Anders Petrelius and Eva Lydahl, St. Erik Eye Hospital, Stockholm, Sweden; Anne-Sophie Strömbeck and Virpi Khalifeh Barke, Department of Ohthalmology, Uppsala University, Uppsala, Sweden. 

⁵  See the Appendix for the names and affiliations of the members of the Swedish Society of Corneal Surgeons who participated in this study.

 Supported by grants from the Gunnar and Märta Bergendahl Foundation (US), and funding from the County Council of Östergötland (PF).

 Submitted for publication September 22, 2008; revised November 27, 2008; accepted March 26, 2009.

 Disclosure: N. Lagali, None; U. Stenevi, None; M. Claesson, None; P. Fagerholm, None; C. Hanson, None; B. Weijdegård, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Graft Recipient Details and Results of FISH Analysis of Cells in Explanted Corneal Buttons

Recipient	Sex	Age* (y)	Indication for Primary Transplant	Graft Age, † (mo)	Surviving Donor Cells (%)			Compensated Grafts, ‡
											Keratocytes	Endothelial
1	F	81	Edema	24	12	0
2	M	88	Band keratopathy	48	59	100
3	F	85	Keratoconus	264	13	0
4	M	41	Edema	6	71
5	F	80	Endothelial dystrophy	55	77	24
6	M	73	Edema	18	5
7	F	78	Edema	3	91	100	Infection
8	F	62	Endothelial dystrophy	52	68	57
9	F	81	Endothelial dystrophy	134	66	61
10	M	62	Keratoconus	90	68	79
11	F	53	Keratoconus	72	53
12	M	61	Keratoconus	28	21	0
13	F	59	Endothelial dystrophy	156	68
14	M	58	Keratoconus	17	83	0
15	M	46	Keratoconus	45	15	0	Astigmatism
16	F	77	Keratitis	6	84	75
17	M	29	Keratoconus	13	38	0
18	M	75	PBK	26	95
19	F	77	Lattice dystrophy	193	78	24
20	F	72	PBK	21	45
21	M	70	Edema	37	58
22	F	59	Endothelial dystrophy	36	93	41
23	F	58	Keratoconus	156	86	0	Astigmatism
24	M	87	PBK	84	58	17
25	M	50	Keratoconus	216	21	6	Astigmatism
26	M	42	Keratoconus	192	89	77	Astigmatism
27	M	44	Herpetic macula	111	15	0	Scar
28	M	39	Keratoconus	120	74	52
29	M	60	Keratoconus	166	66	17
30	M	85	Edema	60	16
31	M	55	Keratoconus	196	8	95
32	F	70	Endothelial dystrophy	24	81
33	M	83	Edema	27	34	18
34	M	77	Edema	47	23	33	Scar
35	F	80	PBK	28	13
36	F	71	Edema	28	7
37	F	71	Edema	27	79
38	F	82	Edema	36	33	69
39	M	81	Keratoconus	210	52	18
40	F	77	PBK	18	95	84	Infection
41	F	77	PBK	27	6
42	F	58	Keratoconus	72	93	70
43	M	31	Keratoconus	103	34	67
44	F	73	Endothelial dystrophy	216	84	42
45	M	45	Keratoconus	240	4
46	M	45	Keratoconus	216	58	31
47	F	61	PBK	53	72
48	M	62	Keratoconus	240	13	8	Scar
49	M	56	Keratoconus	384	65	62	Astigmatism
50	F	53	Keratoconus	96	24
51	M	53	Edema	20	64
52	M	71	Keratoconus	360	9	0

 PBK, pseudophakic bullous keratopathy.

*  Age at time of explantation.

†  Time interval between initial transplantation and explantation.

‡  Reason for explantation of well-compensated, nonedematous grafts. Infections were suture related, astigmatism was intractable in totally transparent grafts, scars were small and localized to the visual axis in otherwise transparent grafts.

Figure 1.

 

Typical fluorescence microscope images used for FISH analysis of corneal sections. (A) Epithelial (bottom) and stromal (top) cells in a female donor corneal button removed from a male recipient. All epithelial cells with two distinct signals had one red and one green signal per cell. Keratocytes had either one red and one green signal (arrow) or two red signals (arrowhead) per cell. (B) Endothelial cells at the posterior surface of a male donor corneal button removed from a female recipient. Endothelial cells with one red and one green signal (arrow) or two red signals (arrowhead) per cell were observed. Scale bar: (A) 50 μm, 20× objective; (B) 10 μm; 100× objective.

Figure 2.

 

Donor keratocytes as a proportion of total keratocytes counted in each of 52 corneal grafts, plotted against graft age (time that the graft remained implanted in the recipient). (○) Subset of eight transparent grafts removed due to reasons other than endothelial decompensation. Donor keratocyte survival did not correlate with graft age.

Figure 3.

 

Donor endothelial cells as a proportion of total endothelial cells counted in each of 35 corneal grafts, plotted against graft age. (○) Subset of eight transparent grafts removed due to reasons other than endothelial decompensation. No donor endothelial cells were found in nine grafts, and no recipient endothelial cells were found in two grafts. Donor endothelial cell survival did not correlate with graft age.