Highlights
-
•
Patients with vitreous hemorrhage at the time of their penetrating keratoplasty have over a six-fold increased odds of graft failure.
-
•
Patients with silicone oil tamponade after their penetrating keratoplasty have over a three-fold increased odds of graft failure.
-
•
Patients with complex eyes that require penetrating keratoplasty as well as posterior segment intervention may benefit from targeted counseling based on these findings.
PURPOSE
To analyze the relationship between intraoperative and postpenetrating keratoplasty (PK) posterior segment variables and PK graft survival.
DESIGN
Retrospective clinical cohort study.
SUBJECTS
Patients undergoing PK between May 1, 2007 and September 1, 2018 at a single tertiary center.
METHODS
Chart review for PKs performed was conducted, and the first PK completed at the institution for each patient was included for analysis. Data collected included demographics, medical and ocular history, preoperative and intraoperative findings, and intraoperative and postoperative posterior segment factors (pars plana vitrectomy [PPV], endolaser, retinal detachment [RD], and vitreous hemorrhage [VH]). After univariable analysis, variables were selected for multivariable Cox regression analysis.
MAIN OUTCOME MEASURE
Graft failure, defined as irreversible and visually significant corneal edema, haze, or scarring.
RESULTS
Eight hundred and thirty-five eyes of 835 patients were included. Mean age was 57.1 ± 22.0 (range: 0-100) years, and mean time from PK to final follow-up or graft failure was 3.2 ± 2.9 (range: 0.01-16.1) years. Graft failure occurred in 35.0% of cases with a mean onset of 1.9 ± 2.0 (range: 0.04-11.4) years after PK. After multivariable analysis, 9 variables had significant associations with failure. Two posterior segment variables were significant: intraoperative VH at the time of PK (hazard ratio [HR] 6.6, 95% confidence interval [CI] 1.6-27.7, P =.010) and silicone oil (SO) tamponade after the PK (HR 3.2, 95% CI 1.4-7.4, P =.007).
CONCLUSIONS
Graft failure is a serious complication of PK. VH at the time of the PK and SO tamponade after the PK were associated with graft failure. In complex eyes that are undergoing PK grafts and that may also require posterior segment interventions, these findings may guide patient counseling and discussion of graft prognosis.
INTRODUCTION
C orneal transplantation is the most common allogenic transplant performed worldwide. Although the use of Descemet’s membrane endothelial keratoplasty (DMEK) and Descemet’s stripping automated endothelial keratoplasty (DSAEK) has increased for the treatment of corneal endothelial disorders, full-thickness replacement via penetrating keratoplasty (PK) remains a common method of transplantation for pathology involving the stroma or the entire cornea. Commonly reported indications for PK include keratoconus, postinfectious corneal scarring or thinning, and pseudophakic bullous keratopathy. , Prior large studies have reported graft failure rates of up to 40% following PK and identified risk factors. ,,,,,,, One study developed a nomogram to predict PK graft prognosis based on multiple independent variables.
The corneal endothelium pumps fluid out of the stroma to maintain transparency and maintain graft viability. Endothelial cell density, however, often decreases after corneal transplantation. , Events that occur in the posterior chamber have the potential to further impair corneal endothelial cell function. ,, Some studies have reported outcomes of PK in combination with posterior segment procedures; ,, however, few studies have analyzed the potential role of posterior segment pathologies or procedures that occurred either during or after the PK. The objective of this study was to investigate the association of posterior segment risk factors, specifically pars plana vitrectomy (PPV), endolaser, retinal detachment (RD), and presence of vitreous hemorrhage (VH) during or after PK to identify associations with PK graft failure. Such data will aide in patient counseling and offer prognostic details for complex eyes that are undergoing PK grafts and that may also require posterior segment interventions.
MATERIALS AND METHODS
This retrospective cohort study was approved by the Institutional Review Board at Wills Eye Hospital. All work was performed in accordance with the Health Insurance Portability and Accountability Act of 1996 and adhered to tenets of the Declaration of Helsinki. The electronic medical records (EMR) at Wills Eye Hospital were queried for all PKs performed between May 1, 2007 and September 1, 2018 using Current Procedural Terminology codes. Surgical technique and postoperative care for PKs have been previously described. , The first PK was included for each patient; sequential PKs, including in contralateral eyes, were excluded to eliminate redundant risk factors. All charts were reviewed in the EMR of the Cornea Service and the Retina Service of Wills Eye Hospital for diagnoses and surgeries that occurred either on the same day as the PK or after transplantation until graft failure or last follow-up.
Collected data included patient demographics, medical and ocular history, preoperative and intraoperative clinical variables, and posterior segment procedures or conditions occurring during or after transplantation ( Table 1 ). The primary outcome was graft failure, which was defined as irreversible and visually significant corneal edema, haze, or scarring. Systemic autoimmune disease, anterior uveitis/endophthalmitis, glaucoma in either eye, prior anterior segment surgery, and dry eye were defined similarly to our center’s prior study. If multiple of the same posterior segment variables were present in the same PK eye, only the first was included to reduce repetition of risk factors and possible overestimation of their effect on graft failure. Events that occurred after graft failure were excluded. Intravitreal injections were not included due inconsistent documentation during the transition from paper charts to EMR. Retinal tamponades were considered postoperative variables, even if placed intraoperatively, as these remained in the eye postoperatively.
TABLE 1
Demographic and Clinical Characteristics (N = 835).
| Variable | N (%) | N Failed (%) | Variable | N (%) | N Failed (%) |
|---|---|---|---|---|---|
| Demographics | Quadrants of corneal neovascularization | ||||
| Age at transplant | 1 | 18 (2.2) | 4 (22.2) | ||
| Sex (female) | 421 (50.4) | 156 (37.1) | 2 | 53 (6.3) | 20 (37.7) |
| Operative PK eye (right) | 418 (50.1) | 149 (35.6) | 3 | 11 (1.3) | 2 (18.2) |
| Medical History | 4 | 30 (3.6) | 14 (46.7) | ||
| Diabetes | 114 (13.7) | 46 (40.4) | Dry eye | 288 (34.5) | 90 (31.3) |
| Systemic autoimmune disease | 29 (3.5) | 16 (55.2) | Anterior synechiae | 123 (14.7) | 68 (55.3) |
| Smoking | 152 (18.2) | 64 (42.1) | Posterior synechiae | 37 (4.4) | 14 (37.8) |
| Ocular History | Preoperative lens status | ||||
| Anterior uveitis/endophthalmitis | 36 (4.3) | 24 (66.7) | Phakic | 435 (52.1) | 100 (23.0) |
| HSV/VZV | 115 (13.8) | 50 (43.5) | Aphakic | 68 (8.1) | 44 (64.7) |
| Trauma | 61 (7.3) | 31 (50.8) | PCIOL | 276 (33.1) | 124 (44.9) |
| Microbial keratitis with scar | 73 (8.7) | 21 (28.8) | ACIOL | 56 (6.7) | 24 (42.9) |
| Chemical burn | 4 (0.5) | 2 (50.0) | Intraoperative | ||
| Glaucoma in either eye | 381 (45.6) | 195 (51.2) | Lens status at end of surgery | ||
| Glaucoma drops in either eye | 323 (38.7) | 171 (52.9) | Phakic | 322 (38.6) | 59 (18.3) |
| Prior PK | 221 (26.5) | 107 (48.4) | Aphakic | 74 (8.9) | 52 (70.3) |
| Number of prior PK | PCIOL | 369 (44.2) | 149 (40.4) | ||
| 1 | 171 (20.5) | 78 (45.6) | ACIOL | 70 (8.4) | 32 (45.7) |
| 2 | 38 (4.6) | 20 (52.6) | Pars plana vitrectomy concurrent with PK | 19 (2.3) | 15 (78.9) |
| 8 (1.0) | 6 (75.0) | Retinal endolaser | 10 (1.2) | 9 (90.0) | |
| 4 | 4 (0.5) | 3 (75.0) | Retinal detachment at time of PK | 13 (1.6) | 12 (92.3) |
| Prior EK | 33 (4.0) | 11 (33.3) | Vitreous hemorrhage at time of PK | 3 (0.4) | 3 (100.0) |
| Contralateral prior PK | 137 (16.4) | 44 (32.1) | Postoperative | ||
| Number of contralateral prior PK | Pars plana vitrectomy after PK | 18 (2.2) | 9 (50.0) | ||
| 1 | 113 (13.5) | 36 (31.9) | Retinal endolaser | 7 (0.8) | 3 (42.9) |
| 2 | 20 (2.4) | 5 (25.0) | Retinal detachment after PK | 13 (1.6) | 9 (69.2) |
| 3 | 4 (0.5) | 3 (75.0) | Vitreous hemorrhage after PK | 8 (1.0) | 6 (75.0) |
| Contralateral prior EK | 10 (1.2) | 2 (20.0) | Silicone oil tamponade | 16 (1.9) | 14 (87.5) |
| Prior anterior segment surgery | 507 (60.7) | 233 (46.0) | C3F8 tamponade | 3 (0.4) | 2 (66.7) |
| Prior glaucoma surgery | 149 (17.8) | 94 (63.1) | Air tamponade | 24 (2.9) | 13 (54.2) |
| Preoperative Exam | |||||
| Active microbial infection | 94 (11.3) | 54 (57.4) | Total | 835 (100) | 292 (35.0) |
| Corneal neovascularization | 112 (13.4) | 40 (35.7) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
