Corneal diseases represent the second leading cause of blindness globally. Keratoplasty is the major surgical procedure for visual restoration of corneal blindness. Corneal wound dehiscence (WD) is not an uncommon complication after keratoplasty. Although its incidence is relatively low, compared with other complications [
From January 2016 to June 2017, 3017 keratoplasties were performed in Shandong Eye Hospital, including 1385 PK surgeries, 1632 LK surgeries, and 75 endothelial transplantations. The patients were 1988 males and 1029 females. We retrospectively reviewed the medical records of these patients and recorded the characteristics, risk factors, and outcomes of corneal WD in patients with different surgical approaches. Data collected included patient age and gender, indication for keratoplasty, surgical procedures, duration between keratoplasty and WD, causative events for WD, size of dehiscence, treatment procedures, and vision outcomes after surgical repair.
All patients with WD were given eye shields or glasses to protect the eyes before treated with emergency corneal graft surgery. Patients who did not have eye content exposure or other serious complications underwent original corneal graft repair under topical anesthesia or general anesthesia. In the eyes with iris or vitreous prolapse, corneal graft rejoint surgery was performed, and complicated fundus surgery was combined if needed. WD was sutured using 10-0 nylon sutures. Postoperatively, systemic and intravenous antibiotics were administered, as well as topical antirejection drugs. All data analyses were performed using SPSS statistical software (version 17.0, SPSS, Inc, Chicago, Illinois, USA). Quantitative data are presented as the mean ± standard deviation (range). A value of
Thirty-one eyes from 30 patients (1.0%) suffered WD after keratoplasty, including 26 males (86.7%) and 4 females (13.3%). The age span of the patients with WD was between 12 and 78 years old with the mean age being 44.6 ± 18.3 years old (Figure
Age of patients with wound dehiscence.
The corneal graft WD occurred in 23 eyes (23/31, 74.2%) after PK and 8 eyes (8/31, 25.1%) after LK. The incidence of WD after PK and LK was 1.66% and 0.49%, respectively (
The indications of PK and LK.
The time interval between keratoplasty and occurrence of WD.
As shown in Table
Characteristics of wound dehiscence following keratoplasty.
Case | Indication for keratoplasty | Type of keratoplasty | Age at the time of trauma (years) | Cause of trauma | Interval between trauma and keratoplasty (months) | Final visual acuity |
---|---|---|---|---|---|---|
1 | PBK | PK | 40 | Unknown | 6 | FC/BE |
2 | HSK | PK | 51 | Struck by iron drill | 13 | FC/20 cm |
3 | KCN | PK | 17 | Finger poke | 76 | LP |
4 | FK | PK | 58 | Struck by own hand | 2 | 20/400 |
5 | HSK | PK | 48 | Spontaneous | 1 | 20/1000 |
6 | PBK | PK | 75 | Struck by wooden stick | 29 | 20/167 |
7 | Bacterial keratitis | PK | 71 | Struck by wooden stick | 13 | LP |
8 | HSK | PK | 62 | Unknown | 99 | HM/10 cm |
9 | Corneal endothelial decompensation | PK | 42 | Spontaneous | 126 | LP |
10 | PBK | PK | 78 | Spontaneous | 52 | 20/500 |
11 | FK | PK | 48 | Struck by desk | 61 | 20/1000 |
12 | FK | PK | 54 | Unknown | 12 | HM/BE |
13 | FK | PK | 45 | Struck by rebar | 22 | 20/400 |
14 | FK | PK | 12 | Struck by book | 27 | HM/50 cm |
15 | KCN | PK | 19 | Struck by phone | 34 | 20/67 |
16 | KCN | PK | 23 | Struck by basketball | 21 | 20/200 |
17 | FK | PK | 49 | Struck by shoes | 1.5 | 20/133 |
18 | HSK | PK | 60 | Struck by wooden stick | 84 | 20/167 |
19 | FK | PK | 43 | Struck by cabbage | 48 | FC/BE |
20 | FK | PK | 29 | Fall | 56 | HM/40 cm |
21 | Interstitial keratitis | PK | 53 | Punch | 84 | 20/67 |
22 | Corneal perforation | PK | 50 | Struck by door | 84 | 20/133 |
23 | Interstitial keratitis | PK | 53 | Punch | 84 | 20/40 |
24 | Ocular chemical injury | LK | 46 | Struck by cages | 204 | HM/BE |
25 | FK | LK | 39 | Punch | 24 | 20/200 |
26 | KCN | LK | 16 | Struck by elbow | 28 | Unknown |
27 | FK | LK | 72 | Struck by own hand | 72 | 20/50 |
28 | Interstitial keratitis | LK | 47 | Unknown | 11 | HM/30 cm |
29 | KCN | LK | 48 | Unknown | 5 | 20/400 |
30 | KCN | LK | 16 | Spontaneous | 31 | 20/80 |
31 | KCN | LK | 20 | Unknown | 12 | 20/67 |
PBK, pseudophakic bullous keratopathy; HSK, herpes simplex virus; KCN, keratoconus; FK, fungal keratitis; PK, penetrating keratoplasty; LK, lamellar keratoplasty; HM, hand moving; FC, finger counting; BE, before eyes; LP, light perception.
Slit lamp examination showed that the corneal fissure was located in the corneal graft-host interface. Nearly one-third (10/31) of the eyes had sutures in place after trauma. 37.5% of the eyes with LK had sutures, while 30.4% of the eyes with PK had sutures in place. The suture technique used in the keratoplasty was interrupted suture. The mean range of dehiscence was 5.5 o’clock in the eyes with sutures, while 5.2 o’clock in the eyes without sutures. There is no significant difference in wound dehiscence (
Four patients after PK and 2 after LK had wound disruption of 1 to 3 clock hours. Seven after PK and 4 after LK had wound disruption of 4 to 6 clock hours. Nine after PK and 2 after LK had wound disruption of 7 to 9 clock hours. One after PK had disruption of 10–12 clock hours. The wound dehiscence encompassed the inferior temporal quadrant in 4 eyes (26.7%), inferior nasal quadrant in 6 eyes (40.0%), superior nasal quadrant in 9 eyes (60.0%), and superior temporal quadrant in 10 eyes (66.7%). The wound dehiscence with 180° or more occurred in 14 eyes (48.3%) with 12 eyes in PK and 2 eyes in LK. And the incidence of extensive wound dehiscence is not different between PK and LK (
With the increase in the range of corneal WD, the degree of eye prolapses increased. Accompanied complications included iris prolapse in 5 eyes (16.1%), lens expulsion or dislocation in 15 eyes (48.4%), and extrusion of vitreous in 11 eyes (35.5%). In the eyes treated by PK, the complications were iris prolapse in 4 eyes, lens expulsion or dislocation in 15 eyes, and extrusion of vitreous in 10 eyes. In the eyes treated by LK, the complications were iris prolapse in 1 eye and extrusion of vitreous in 1 eye. The lens in one eye and the vitreous in the other eyes were not seen clearly. The extrusion of the lens and vitreous mainly occurred in the patients with an extent of wound disruption
The duration between the occurrence of corneal graft dehiscence and therapy was 2 to 72 hours. Among 31 eyes of 30 patients, 31 eyes, including 22 eyes after PK and 8 eyes after LK, just had the graft repaired, and only 1 eye after PK was treated with combined anterior chamber angioplasty surgery because of flat anterior chamber.
Final visual acuity was 20/200 or better in 12 eyes (40%), better than hand motions (HM) to 20/200 in 11 eyes (36.7%), HM to light perception (LP) in 7 eyes (23.3%), and unknown in one eye. In the follow-up period, BCVA was improved in 19 eyes (65.5%), including 16 eyes with PK and 3 eyes with LK, unchanged in 9 eyes (31.0%), including 6 eyes with PK and 3 eyes with LK, and decreased in one eye with LK (3.5%). Patients after treatment of LK achieved better final visual acuity than those after PK, but the final visual acuity and the recovery of visual acuity were of no statistical significance (
The cornea never regains the original tensile strength after keratoplasty [
It was reported that the incidence of WD was related to age. Older people were found to be more likely to develop graft WD [
WD after keratoplasty has been divided into traumatic WD and spontaneous WD. In our study, the incidence of traumatic WD was 87.1%, but we also needed to notice that there were some cases with no obvious causes. Long-term using of topical corticosteroids could increase the risk of corneal WD after the removal of sutures [
It was reported that corneal WD mostly occurred within two years. The mean interval between keratoplasty and WD was 45.9 months in our study, and dehiscence occurred during the first 4 years in 61.3% of the eyes. The longest duration between keratoplasty and WD occurrence in China was 9 years [
In our hospital, the sutures were removed within 1.5 years. Therefore, the sutures were in place in the eyes in which WD happened within 1.5 years. To our surprise, we found that WD without sutures did not lead to more extensive WD compared with those with sutures. So we think that the remaining sutures did not affect WD. The extend of graft dehiscence attributed to the trauma after WD. This result was not consistent with other reports. We think the reason may be that our sample is too small. Meyer found that leaving sutures may maintain the integrity of the graft-host junction and dehiscence with sutures led to less dehiscence [
In the current study, the type of operation was found to be an important factor of the occurrence of WD. The incidence of WD was 1.66% after PK and 0.49% after LK (
WD may result in many serious ocular complications including iris prolapse, crystalline or intraocular lens expulsion or dislocation, and extrusion of vitreous. As previously reported, lens expulsion or dislocation was associated with poor prognosis and the final visual acuity. We noticed that PK patients tended to suffer more severe complications from WD. The reason may be that the cornea still remained a part of the autologous corneal tissue after LK and was protected with the help of full thickness of the Descemet membrane [
Once WD occurs, the degree of injury would directly affect the patient prognosis. It was reported that in patients with poor prognosis after injury, only 1/3 to 1/2 of patients had visual acuity of 20/200 [
The location of wound dehiscence.
In conclusion, WD is a risk factor for patients undergoing corneal transplant. Compared with LK, PK seems to be more prone to result in wound dehiscence. The WD after LK may be less severe. The visual acuity after treatment of WD can be worse in the eyes with PK than LK. To reduce the incidence of WD after corneal transplantation, the patient’s condition needs to be comprehensively analyzed before selecting appropriate surgical approaches, regular postoperative follow-up is important, and the protective awareness of the patient and family members should be improved.
The authors have no conflicts of interests to declare.
This work was supported by the National Natural Science Foundation of China (81370989, 81570821, and 81530027), Science and Technology Development Program of Shandong Province (2016GSF201182), Taishan Scholar Program (20081148), the Natural Science Foundation of Shandong Province (ZR2015YL026), and Innovation Project of Shandong Academy of Medical Sciences.