To investigate the incidence and outcome of cornea transplant rejection following endothelial keratoplasty (EK) and penetrating keratoplasty (PK) for Fuchs endothelial dystrophy (FED) and pseudophakic bullous keratopathy (PBK).
Multicenter cohort study.
Patients registered on the United Kingdom Transplant Registry (UKTR) who had an EK or PK for FED or PBK between April 1, 2005 and March 31, 2011 were included. Data were collected from UKTR forms at 1 and 2 years. Postoperative steroid use varies between surgeons and cannot be captured in this reporting system. Rejection events were identified as those recorded as endothelial rejection.
A total of 3486 corneal transplants were undertaken: 1973 for FED, 1513 for PBK. For FED, 2-year rejection-free survival was 93% (95% confidence interval [CI] 90%–94%) for PK and 94% (95% CI 92%–96%) for EK ( P = .3). In transplants that had a rejection episode, 50% of PKs (17) and 60% of EKs (15) subsequently failed. For PBK, 2-year rejection-free survival for PK was 88% (95% CI 86%–90%) and 90% (95% CI 86%–92%) for EK ( P = .6). In transplants that had a rejection episode, 85% of PKs (41) and 76% of EKs (22) subsequently failed. Inflammation (ie, conjunctival injection, presence of keratic precipitates and intraocular signs) at the time of surgery for patients with FED was significant for developing rejection: 3.5 times greater compared with those with no inflammation ( P = .02).
There is no significant difference in rejection-free survival between EK and PK for FED or PBK. The presence of inflammation is an important risk factor, and attention to its control before and following surgery is important.
Almost half of all corneal transplants in the United Kingdom (UK) are performed for visual loss due to endothelial dysfunction (UK Transplant Registry, OTAG17). Over the past 15 years, the technique for endothelial keratoplasty (EK) has developed and has been continually refined, such that EK (known as deep lamellar endothelial keratoplasty [DLEK], Descemet stripping endothelial keratoplasty [DSEK], Descemet stripping automated endothelial keratoplasty (DSAEK), and Descemet membrane endothelial keratoplasty [DMEK]) is now proposed as the surgical procedure of choice for patients with endothelial dysfunction.
Corneal transplant rejection, however, remains an important factor affecting the outcome of all forms of corneal transplantation. In comparison to penetrating keratoplasty (PK), there is a theoretically reduced risk of rejection following EK owing to at least 4 factors : (1) the transplanted tissue is placed in the anterior chamber with no apparent exposure to the antigen-presenting cells in anterior recipient cornea and antibodies of ocular surface; (2) there are no transplant suture–related rejection episodes; (3) in cases of a vascularized host cornea, the absence of direct contact between the host stromal vessels and the transplanted tissue reduces the impact of vascularization on immune cell trafficking to and from donor tissue; and (4) there is reduced immunogenicity of the donor tissue owing to absence of epithelium and less stromal tissue. There have been many studies investigating the incidence of rejection following PK, but the reported incidence of rejection and subsequent graft failure following EK varies very widely. Based on data amalgamated from 4 centers, Allan and associates reported a significantly lower incidence of rejection following EK (7.5%) than PK (13%). The same authors also noted a trend ( P = .06) for graft failure to be less frequent after rejection in EK (6.7%) than in PK (28.3%). The 2 groups, however, had different treatment regimens and were from different centers; that is, all the patients who had PK were from the Swedish Corneal Transplant Registry. Nanavaty and Shortt, in their Cochrane Review, noted that for Fuchs endothelial dystrophy (FED), EK was associated with a lower rejection rate (2%–4%) compared to PK (5.3%–23.2%), although the mean follow-up was considerably longer for the PK than the EK group. In contrast, Lee and associates reviewed 34 peer-reviewed articles and concluded that the mean rejection rate following EK was 10% and was similar to that following PK, but with a high variation of between 0% and 45.5%. These data are supported by various individual studies, some of which were included in the review by Lee and associates.
A summary of rejection rates following EK for studies published since 2007 is shown in Table 1 . This suggests that based on a median follow-up period of 24 months (mean 25.5, range 12–84, standard deviation [SD] 20.4), the median rejection rate for EK is 6.7% (mean 7.4%, range 2.5–14, SD 3.5). Graft survival following EK has been recently reported to be significantly lower than for PK in the United Kingdom, in a report investigating any correlation between graft survival and refractive outcomes with surgeon/center experience, without investigating the effect of rejection on graft survival. The Australian Corneal Graft Registry also reported similar results. Given the importance of graft rejection as a risk factor for graft failure, we investigated the incidence and outcome of corneal transplant rejection following PK and EK for FED or pseudophakic bullous keratopathy (PBK).
|Author||Year||Eyes||Follow-up (Months)||Rejection (%)|
|Pedersen et al||2015||78||48||4|
|Khor et al||2013||78||12||2.5|
|Ezon et al||2013||122||29||5|
|Wu et al||2012||353||24||14|
|Anshu et al||2012||598||13||9|
|Li et al||2012||615||–||7.3|
|Price et al||2009||598||24||12|
|Price et al||2008||40||12||5|
|Suh et al||2008||118||–||6|
|Heidemann et al||2008||20||12||5|
|Terry et al||2008||80||12||7.5|
|Price et al||2008||263||24||0.76|
|Van Dooren et al||2007||15||84||6.7|
|Covert et al||2007||7||13.3||14|
|Allan et al||2007||199||24||7.5|
|Total: 3184||Median: 24 (mean 25.5, range 12–84, SD 20.4)||Median: 6.7 (mean 7.4, range 2.5–14, SD 3.5)|
In the UK, data on patients undergoing corneal transplantation who are registered with National Health Service Blood and Transplant (NHSBT) are stored in the UK Transplant Registry (UKTR). This information is collected at the time of surgery and at 1, 2, and 5 years following transplantation. Events occurring in the intervening period are captured at the next follow-up form (ie, an event at 6 months follow-up will be captured at the 1-year follow-up form). Completion of the transplant record and follow-up forms is a requirement for all transplants registered with NHSBT.
This is a multicenter cohort study of all patients registered on the UKTR of NHSBT who had undergone an EK or PK as a first graft for FED and PBK in the period between April 1, 2005 and March 31, 2011. The described research methods and analysis plan adhered to the tenets of the Declaration of Helsinki, and Institutional Review Board of NHSBT approval was obtained. Data were collected from transplant records and follow-up forms at 1 and 2 years postoperatively; these data included any events that may have occurred in the interim period between each annual visit. Clinical information collected included preoperative data (determined by the operating surgeon at the time of operation, including the presence of any inflammation [ie, conjunctival injection, the presence of keratic precipitates, intraocular inflammation signs], any ocular surface disease [eg, allergic or dry eye disease, meibomian gland dysfunction, epithelial instability, corneal ulceration], glaucoma, and quadrants of corneal vascularization), perioperative events (ie, donor and recipient characteristics, complications), and postoperative outcome (including visual acuity, complications, medications, refractive data, rejection episodes and type of rejection, graft survival, and reasons for graft failure). Rejection events were identified as those recorded as endothelial rejection (corneal edema associated with anterior chamber activity, with keratic precipitates ± an endothelial rejection line). Standard management of all corneal transplants in the UK include the use of a topical steroid, initially at a high dose, before being gradually tapered. The dose and frequency of administration, however, varies between surgeons and centers and this information cannot be captured on the transplant record and follow-up forms.
Statistical analyses were performed using SAS v9.1 software (SAS Institute Inc, Cary, North Carolina, USA). χ 2 tests were used to investigate differences in categorical variables. Kaplan-Meier survival curves were used to compare rejection-free graft survival between PK and EK. A Cox regression model was fitted to compare the relative risk of rejection following EK and PK, adjusting for the following possible risk factors: vascularization and inflammation.
A Medline literature search using the terms “endothelial keratoplasty,” “endothelial keratoplasty rejection,” and “endothelial keratoplasty outcomes” was undertaken to identify publications and construct a meta-analysis table of rejection rates in EK ( Table 1 ).
A total of 3486 corneal transplants were undertaken. Of these, 1973 transplants were for FED (1002 PKs and 971 EKs) and 1513 for PBK (959 PKs and 554 EKs). Two-year follow-up data were available for, respectively, 90% (902/1002) and 80.5% (782/971) of PKs and EKs undertaken for FED, and 87.3% (837/959) and 79.2% (439/554) of PKs and EKs undertaken for PBK. Registry data did not, for the period of study, distinguish between the type of EK (ie, DMEK and DSAEK), but during the period studied (2005–2011) very few, if any, DMEKs were undertaken.
The 1- and 2-year rejection-free survival rates for transplants undertaken for FED and PBK were similar for PK and EK, respectively ( Figures 1 and 2 , Table 2 ). For FED, the 1-year rejection-free survival rate for PK was 96% (95% confidence interval [CI] 94%–98%) compared with 97% (95% CI 96%–98%) for EK ( P = .32), with a rejection episode reported in 34 PKs and 25 EKs. Of those transplants for which a rejection episode was recorded, 50% of PKs (17) and 60% of EKs (15) subsequently failed during the recorded follow-up period. For PBK, the 1-year rejection-free survival rate for PK was 94% (95% CI 92%–96%), compared with 94% (95% CI 92%–96%) for EK ( P = .81), with a rejection episode reported in 48 PKs and 28 EKs. Of those transplants that had a rejection episode, 85% of PKs (41) and 76% of EKs (22) subsequently failed. A similar pattern was observed for both procedures for both FED and PBK at 2 years ( Table 2 ).
|1 Year||2 Years|
|Rejection-Free Survival||Failures Following Rejection||Rejection-Free Survival||Failures Following Rejection|
|FED||PK||96% (95% CI 94%–98%)||17/34||93% (95% CI 90%–94%)||32/61|
|EK||97% (95% CI 96%–98%)||15/25||94% (95% CI 92%–96%)||26/45|
|P = .32||P = .30|
|PBK||PK||94% (95% CI 92%–96%)||41/48||88% (95% CI 86%–90%)||69/90|
|EK||94% (95% CI 92%–96%)||22/28||90% (95% CI 86%–92%)||32/42|
|P = .81||P = .60|