Methods

In a retrospective cohort study, data from the UK Transplant Registry were provided by and analyzed in collaboration with NHS Blood and Transplant, which provided institutional board approval. In the UK Transplant Registry, blanket informed consent is obtained at the time of corneal transplantation from all patients, allowing the use of registry data for scientific purposes. The study cohort comprised patients over 18 years in the UK with FED, PBK, and KC (which are the main indications for corneal transplantation ) who had undergone either a PK, EK, or DALK between April 1, 1999 and March 31, 2012. The bilateral cohort included all patients who had received a first transplant in both eyes, where the indication was the same in both eyes and the transplant in the second eye occurred within the follow-up period for the transplant in the first eye (maximum of 5 years) and before graft failure in the first eye. A control cohort included all patients who underwent unilateral first corneal transplant for the same indication during the same period. Patients who underwent a transplant in the second eye beyond the follow-up period for the first eye or after graft failure in the first eye were also included in the control cohort. Patient and outcome data were collected at the time of the transplant and then at 1 year, 2 years, and 5 years post-transplant. All data were obtained from the UK Transplant Registry for corneas supplied through the UK corneal transplant service (CTS) eye banks.

Analysis

Recipient and transplant characteristics of the bilateral cohort were summarized as counts and percentages. For the bilateral cohort, the median time between transplants in the first and second eye was defined as the inter-transplant time (ITT) and compared across indications using the Kruskal-Wallis test. Kaplan-Meier estimates of first and second eye graft survival from time of second eye graft were compared across ITT categories (<1 year, 1–2 years, >2 years) for each indication using the log-rank test. The association between ITT and first and second eye graft survival from time of second eye graft was further investigated using Cox regression. Nonlinear associations between ITT and first and second eye graft survival were assessed using the likelihood ratio test, comparing a linear form of ITT with the categorical form of ITT. In analyses comparing the bilateral cohort and the control cohort, survival and rejection-free rates for the first eye graft at 5 years post-transplant were estimated using the Kaplan-Meier method. The association between receiving a graft in the second eye and survival and rejection of the first eye graft was investigated using Cox regression. Because it was not known if and when a recipient would receive a transplant in the fellow eye at the time of the transplant in the first eye, a time-dependent indicator variable that changed from 0 to 1 at the time of the transplant in the second eye was used in the model, as shown in Figure 1 . Regression analyses were performed separately for patients with FED, PBK, and KC, adjusting for the type of transplant (PK, DALK, and EK) as appropriate to that indication, low- or high-risk recipient (high risk defined as any ocular surface disease, corneal vascularization, or glaucoma at time of transplant), and, for KC and FED patients, postoperative intraocular surgery (predominantly cataract surgery) in either eye following the transplant prior to graft failure and/or rejection. Postoperative intraocular surgery was included as a time-dependent variable. Analyses were undertaken using SAS/STAT version 9.4 (SAS Institute Inc, Cary, North Carolina, USA).

Schematic of graft survival in the first eye, illustrating a time-dependent Cox regression model. Estimated baseline survivor function before and following transplant in the second (fellow) eye shown. Recipients start on the baseline survivor curve (solid line). If a recipient never receives a transplant in their second eye, they remain on the bottom curve. At the time of the transplant in the second eye, the recipient “jumps” to the top curve, representing graft survival for the bilateral group. The vertical dotted line represents the median time to the transplant in the second eye, that is, the median time when the “jump” would occur. At the time of each graft failure in the cohort, the risk set includes only those patients who have survived to that time. This allows a direct comparison of patients with identical survival who had or had not received a second eye graft at that time. If the second eye graft for a particular recipient occurred after 1 such event, that recipient would be included in the “no second eye graft” category until the time of their second eye graft. The hazard ratio associated with the second eye graft is then interpreted as the hazard of graft failure for a recipient who received a second eye graft compared with one who did not, where the 2 recipients have the same values of all other covariates at that time.

Results

A total of 11 822 patients were included, of whom 2487 had undergone bilateral corneal transplantation within the study period. Of the 2487 patients with bilateral corneal transplants, 1584 (64%) had undergone transplants in both eyes within the follow-up period for the first eye and before graft failure in the first eye; 743 patients (30%) had a functioning graft in the first eye at last follow-up but received a transplant in the second eye beyond the known follow-up period for the first eye; 114 patients (5%) received a transplant in the second eye after having had a repeat transplant in the first eye; and 46 patients (2%) received a transplant in the second eye after graft failure in the first eye, but before or without a repeat transplant in the first eye. Therefore, 1584 patients were defined as bilateral recipients for the purpose of this study. Of these, the proportions receiving the same type of transplant in the second eye as in the first eye were as follows: EK, 98%, PK, 80%, and DALK, 78%. Follow-up data for the transplant in the second eye were missing for 59 patients (3.7%). For these recipients, only data for the transplant in the first eye were used. Recipient and transplant characteristics for the bilateral cohort are shown in Table 1 .

Median ITT was higher for patients with KC (median 734 days, interquartile range [IQR] 503–1110 days) than for those with FED (median 616 days, IQR 378–981 days) and lowest for PBK (median 535 days, IQR 322–823 days) ( P < .0005). We found no evidence of a nonlinear association between ITT and first and second eye graft survival for any indication ( P > .2 in all cases) and hence a continuous linear form of ITT was used in all regression models. Graft survival in the first eye at 3 years from transplant in the second eye was higher for patients with KC (97.9%, 95% CI: 95.6%–99.0%) than for those with FED (92.4%, 95% CI: 89.5%–94.5%) and lowest for patients with PBK (74.7%, 95% CI: 60.1%–84.7%) ( P < .0005). Importantly, however, within each indication first eye graft survival estimates were similar, regardless of ITT (KC P = .16, FED P = .25, PBK P = .17) ( Figure 2 ). After risk adjustment, we found no association between ITT and graft survival in the first eye (KC P = .14, FED P = .59, PBK P = .97) ( Table 2 ). Similar to the first eye, graft survival in the second eye at 3 years from transplant in the second eye was higher for patients with KC (94.6%, 95% CI: 91.8%–96.4%) than for those with FED (88.5%, 95% CI: 86.0%–90.6%) and lowest for those patients with PBK (75.3%, 95% CI: 61.4%–84.8%) ( P < .0005). Again, within each indication, estimates of graft survival in the second eye graft were similar, regardless of ITT (KC P = .80, FED P = .74, PBK P = .36). After risk adjustment, we found no association between ITT and graft survival in the second eye (KC P = .92, FED P = .07, PBK P = .38) ( Table 2 ).

First eye graft survival, from the time of transplant in the second eye, by inter-transplant time (ITT) and indication for the bilateral cohort. Within each indication, graft survival in the first eye was similar for all categories of ITT (keratoconus [KC] P = .16, Fuchs endothelial dystrophy [FED] P = .25, pseudophakic bullous keratopathy [PBK] P = .17).

For all 11 822 patients who received a first eye graft, the overall 5-year graft survival rates were 91% (95% CI: 90%–92%) for KC, 78% (95% CI: 76%–80%) for FED, and 53% (95% CI: 51%–56%) for PBK. The overall 5-year rejection-free rates were 83% (95% CI: 81%–84%) for KC, 87% (95% CI: 85%–88%) for FED, and 77% (95% CI: 75%–79%) for PBK. Analysis of first eye graft outcomes for the whole study cohort indicated that, for patients with FED and KC, the risk of graft failure within 5 years in the first eye for patients who received a transplant in the second eye was half that of patients who did not receive a transplant in the second eye (FED: hazard ratio [HR] 0.47, 95% CI: 0.34–0.64, P < .005; KC: HR 0.50, 95% CI: 0.24–1.02, P = .03) ( Table 3 ). By including an interaction term in each model (type of transplant in the first eye by second eye graft), we found that this effect was independent of the type of transplant in the first eye for KC ( P = .23) but not for FED ( P = .03, EK: HR 0.30, 95% CI: 0.17–0.52; PK: HR 0.61, 95% CI: 0.42–0.88). For patients with PBK we found no association between a transplant in the second eye and graft failure in the first eye, regardless of the type of transplant in the first eye (HR 0.69, 95% CI:0.42–1.14, P = .61). As an additional measure of how well the first eye graft was doing post-transplant, we also included post-transplant corrected logMAR visual acuity in the first eye as a time-varying covariate in each model. We still found that a transplant in the second eye was associated with a reduced risk of graft failure, although the confidence intervals are far wider owing to missing values of logMAR visual acuity (FED: HR 0.69, 95% CI: 0.45–1.04; KC: HR 0.49, 95% CI: 0.20–1.18; PBK: HR 0.76, 95% CI: 0.34–1.71). In addition, we found no association between a transplant in the second eye and the development of a rejection episode in the first eye, regardless of the type of transplant in the first eye (KC P = .48, FED P = .11, PBK P = .85) ( Table 3 ). There were only a few cases of graft failure or rejection within 5 years of transplant where the type of transplant in the second eye was different from that for the first eye (KC: graft failure, 2 cases, graft rejection, 13 cases; FED: graft failure, 6 cases, graft rejection, 9 cases; PBK: graft failure, 1 case, graft rejection, 0 cases), so we were not able to estimate the effect of type of transplant in the second eye.

Table 3

Graft Failure and Rejection

Parameter	Level (Baseline)	Transplants (N)	Events (N)	Hazard Ratio (95% CI) for Graft in the Second Eye Compared With No Graft	P Value
Risk of graft failure in the first eye
KC: Graft in second eye a	No	4104	285	1.00 (–)
	Yes	504	7	0.50 (0.24–1.02)	.03
FED: Graft in second eye a	No	2966	540	1.00 (–)
	Yes	991	47	0.47 (0.34–0.64)	<.01
PBK: Graft in second eye b	No	3168	984	1.00 (–)
	Yes	89	15	0.69 (0.42–1.14)	.13
Risk of rejection in the first eye
KC: Graft in second eye a	No	4104	546	1.00 (–)
	Yes	504	58	0.72 (0.43–1.20)	.19
FED: Graft in second eye a	No	2966	277	1.00 (–)
	Yes	991	78	0.85 (0.59–1.23)	.39
PBK: Graft in second eye b	No	3168	466	1.00 (–)
	Yes	89	11	0.55 (0.20–1.47)	.19

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Resolving Genetic Test Results for the Patient and the Clinician
2. Changing Practice Patterns and Long-term Outcomes of Endothelial Versus Penetrating Keratoplasty: A Prospective Dutch Registry Study
3. Long-term Repeatability and Reproducibility of Phosphene Characteristics in Chronically Implanted Argus II Retinal Prosthesis Subjects
4. Reply

Stay updated, free articles. Join our Telegram channel

Join