Slitlamp photograph of traumatized eye of patient 7. Note the broken suture and dehiscence of lower nasal surgical wound, the expelled crystalline lens in the lower fornix, and vitreous strand adhered to the lens and upper fornix.
Slitlamp photograph of patient 11 after the trauma. Note dehiscence of the surgical wound in the superior half, vitreous in the upper temporal cul-de-sac, graft edema, and Descemet folds. The crystalline lens was expelled from the eye through the keratoplasty wound dehiscence.
Slitlamp photograph of the traumatized eye seen in Figure 2, 2 months following surgical repair that included vitrectomy, scleral fixation of a posterior chamber intraocular lens implant, and resuture of the original corneal graft with running 10-0 monofilament nylon suture.
Rehany U, Rumelt S. Ocular Trauma Following Penetrating KeratoplastyIncidence, Outcome, and Postoperative Recommendations. Arch Ophthalmol. 1998;116(10):1282-1286. doi:10.1001/archopht.116.10.1282
The surgical wound after penetrating keratoplasty is more vulnerable to contusive trauma than the intact cornea.
To assess the incidence of ocular trauma following penetrating keratoplasty, and to evaluate its causative factors, management, and visual outcome.
Tertiary referral facility in a fairly closed population.
Retrospective study assessing ocular injuries of all 559 patients who underwent penetrating keratoplasty in the center between September 1986 and March 1993.
Fourteen (2.5%) of 559 patients who underwent penetrating keratoplasty, over a period of 78 months, suffered surgical wound dehiscence because of contusive ocular trauma. The mean age of the patients (30.6 years) was significantly lower (P<.001) than that of the total number of patients who received transplants (49 years). The interval between transplantation and trauma ranged from 2 weeks to 2 years (mean interval, 6.7 months). In 11 of the 14 patients the trauma occurred prior to removal of sutures. Globe rupture occurred at the donor-recipient interface in all of the patients, accompanied by vitreous and lens loss in 8 patients (57%). In 2 patients (14%) trauma included disruption of retinal tissue resulting in poor visual outcome. The trauma occurred most often at home (in 7 patients). After follow-up periods of 1 to 6 years (mean, 29 months), the corrected visual acuity ranged from 20/120 to 20/20 in 12 patients (86%).
Ocular injury following penetrating keratoplasty is not a rare event. Since corneal graft wound is vulnerable to ocular trauma, we recommend the constant use of protective eyewear and periodically instruct the patients on the long-term vulnerability of the graft wound. Patients should be repeatedly cautioned against high-risk environments and strenuous activity.
PENETRATING keratoplasty exposes the patient to increased risk for wound dehiscence spontaneously or because of blunt trauma,1- 6 since the wound may never gain the tensile strength of a normal intact cornea. The avascular donor-host interface, in most cases, and the prolonged postoperative treatment with topical corticosteroids delay wound healing and may contribute further to the surgical wound weakness.
Poor visual outcome may result from a corneal graft traumatization caused by endothelial cell loss or secondary glaucoma.2- 5 Vision is severely decreased if the posterior segment is involved as a result of vitreous hemorrhage, suprachoroidal hemorrhage, retinal disruption, or prolapse of intraocular tissues,2- 4 and enucleation of the traumatized eye may be required when severe disruption of the ocular tissues occurs.1 Therefore, it is essential to know the incidence of ocular trauma following penetrating keratoplasty that causes wound dehiscence, its temporal relation to the time of surgery, and the mechanism by which it occurs. Knowledge of these data may allow practical postoperative recommendations regarding the use of protective eyewear. The present communication addresses these issues.
The records of all the patients who underwent penetrating keratoplasty at the Western Galilee–Nahariya Medical Center, Nahariya, Israel, between September 1986 and March 1993 were reviewed. Demographic and medical data were evaluated for statistical studies. The records of all patients who received corneal transplants and had traumatic graft wound dehiscence were examined in detail. All of the patients who underwent penetrating keratoplasty during the 78-month period were followed up until March 1994 for this study. Thus, the follow-up period for the total number of patients who received transplants, including those who suffered ocular trauma, was between 12 and 90 months (mean, 29 months).
Between September 1986 and March 1994, 559 penetrating keratoplasties were performed of which 14 eyes of 14 patients (2.5%) suffered surgical wound dehiscence due to contusive ocular trauma. The trauma group included 9 males and 5 females, aged 4 to 64 years. The mean age (30.6 years) of those in the trauma group was significantly lower (P<.001) than that of the total number of patients who received transplants (mean age, 49 years; age range, 4-87 years).
The preoperative diagnoses of the patients who received corneal transplants and who suffered contusive ocular trauma were as follows: keratoconus in 10 patients; herpetic corneal scars in 2; a traumatic corneal scar in 1; and severe irregular astigmatism following refractive corneal surgery in 1 (Table 1). In the trauma group, the corneal allograft diameters ranged from 7.5 to 8.5 mm and the recipient's corneal beds ranged from 7.0 to 8.0 mm. All eyes were phakic. Originally, the corneal graft was secured to its bed by running sutures in 11 patients (78.6%) and interrupted sutures in 3 individuals (21.4%) (Table 1). In the 412 patients (73.7%) who received corneal transplants, a running suture was used. At the conclusion of all the original corneal transplantations, 3 mg of soluble betamethasone acetate and 20 mg of gentamicin sulfate were injected subconjunctivally. Topical 0.1% dexamethasone phosphate was given an average of 4 times daily during the first postoperative month and gradually tapered to 1 drop every day during the first postoperative year. All of the original penetrating keratoplasties and 10 of the surgical repairs after the ocular trauma were performed by the same surgeon (U.R.) after obtaining an informed consent.
All of the patients in this study were injured by blunt trauma. In 7 patients (50%) the trauma occurred at home either by upper extremity (4 patients, finger, fist, elbow) or by inanimate object (3 patients) (Table 2). Two eyes were traumatized accidentally at work by a blunt object; 3, as a result of intentional assaults; 1, due to ocular trauma in a vehicular crash; and 1, in a basketball game. Three of the patients suffered minor trauma due to eye poking in the first postoperative year. Only 2 patients wore protective spectacles at the time of injury. Eight patients suffered injury to their right eye and 6 to their left. The interval between penetrating keratoplasty and the trauma ranged from 2 weeks to 2 years, with a mean of 6.7 months (Table 2).
All patients incurred traumatic dehiscence of the keratoplasty wound, and the sutures, if present, were torn in the area of dehiscence (Figure 1). In 11 patients the traumatic wound dehiscence occurred prior to suture removal, whereas in 3 patients the trauma occurred after suture removal (Figure 2). The wound dehiscence in all 14 resulted in a flat anterior chamber. Iris prolapse was disclosed in 3 patients. Expulsion of the crystalline lens and vitreous loss occurred in 7 patients (50%) (Figure 1 and Figure 2, Table 2). Lens injury accompanied by vitreous loss occurred in 1 patient.
The wound dehiscence was resutured either by a running suture joined to the free ends of the remaining intact running suture (8 eyes) or by multiple interrupted sutures (6 eyes). Anterior vitrectomy was performed in all cases of vitreous loss, and iris repositioning was performed in all cases of prolapsed iris. All patients with lens expulsion were left aphakic except for one, in whom a posterior chamber intraocular lens (IOL) was fixated to the sclera (Table 2, Figure 3).
Thirteen (93%) of the traumatized corneal grafts remained clear, while 1 needed regrafting because of graft edema. There were no episodes of endophthalmitis. Four eyes developed secondary glaucoma. Intraocular pressure was medically controlled in 1 eye. Two eyes underwent trabeculectomy and continuous-wave-contact YAG-laser cyclophotocoagulation was applied in the fourth case to control high intraocular pressure.
The final best-corrected visual acuity after follow-up periods of 12 to 72 months (mean, 29 months) ranged from uncertain light projection to counting fingers at 3 ft in 2 patients (14%). In 3 patients, visual acuity was 20/200 to 20/50, and in 9 patients 20/40 to 20/20. One patient with uncertain light projection developed inoperable traction retinal detachment. Two patients with poor visual outcome suffered from retinal damage and vitreoretinal complications (Table 3).
To our knowledge, few studies on ocular trauma following corneal transplantation have been published,1- 4 none of them pertaining to the incidence of postkeratoplasty trauma, to the age group compared with the average age of the total number of patients who received transplants, and to specific recommendations to decrease the incidence of postkeratoplasty trauma. We have found that ocular injury causing corneal graft dehiscence following penetrating keratoplasty is not a rare event (2.5% of the total population who receive corneal transplants). The trauma group varied in age, sex, ethnic and socioeconomic status, occupation, and lifestyle. The mean age of the trauma group (30.6 years) was significantly lower (P<.001) than that of the total population receiving transplants (49 years). The higher prevalence of ocular trauma in young males was evident (9 of 14 patients), as in other studies,2,3 and may be attributed to the high exposure of younger males to eye injury.7
In the present study, the indication for keratoplasty was keratoconus in 10 (71.4%) of the 14 patients and only in 196 (35%) of the 559 patients who underwent corneal transplantation. The surgical wound in corneal grafts for keratoconus usually remains avascular and heals rather slowly. This may account for the high prevalence of traumatic wound dehiscence. Likewise, these patients are usually younger and at higher risk for trauma.7
The corneal grafts were secured to their beds by running sutures in the 412 patients (73.7%) who received corneal transplants. A similar portion, 11 patients(78.6%), of the trauma group had running sutures. Thus, running sutures are not overrepresented in the trauma group and it is unlikely that they could predispose to traumatic wound dehiscence. In 11 (79%) of the 14 patients, traumatic wound dehiscence occurred prior to suture removal. The high prevalence of traumatic wound dehiscence in the early postoperative period may be explained by 2 factors. The first is the weakness of the corneal wound in the early postoperative period and its high vulnerability to contusive trauma. It has been shown by Gassett and Dohlman8 that approximately 3 months after corneal transplantation in rabbits, the tensile strength of a keratoplasty wound is only 50% that of the normal intact cornea. The donor-host interface is further weakened by the frequent postoperative use of topical steroids to suppress immune response. The second factor may be the change in physical activity of the patients due to the rapid improvement in vision following successful keratoplasty. Young patients in particular may ignore the vulnerability of the operated on eye and allow themselves strenuous activity at home, work, or during sports activities. In 7 patients (50%), the injuries were incurred at home, a further result of the hasty return to normal activity and of the careless attitude toward the sensitivity of the eye recently operated on. The use of modern microsurgical techniques and nonirritating sutures results in a better and faster postoperative recovery, but may also decrease the awareness of the patients to the vulnerability to ocular trauma. Patients at high risk for trauma because of their medical condition and living environment, such as children with mental retardation who live at home (patient 13), are poor candidates for corneal transplantation. When reasonable protection of an eye that undergoes corneal transplantation is not possible, keratoplasty should be avoided even when strongly indicated.
The force of the trauma is usually directly expressed by the severity of the ocular damage caused by it. In 11 of our patients, the trauma was sufficient to break the sutures, and in 6 patients the crystalline lens was expelled. Despite the severe trauma, the final corrected visual outcome was excellent (20/40-20/20) in 9 (64%) of the patients. Similar results have been reported in previous studies,3,4 while others1,2 reported poor outcome. The present study supports studies advocating that visual outcome of ocular trauma depends greatly on the nature and severity of the trauma and the objects causing it. Involvement of the posterior segment results almost always in a poor visual outcome.
Seven of the patients remained aphakic following traumatic expulsion of the crystalline lens (Figure 1 and Figure 2, Table 2). In one patient, a posterior chamber IOL implant was fixated to the sclera through the ciliary sulcus, during the primary surgical repair (Figure 3). The implantation of an IOL during the primary surgical repair has the advantage of avoiding a second surgery in cases where a contact lens is intolerable or contraindicated. Nonetheless, the implantation may also pose a risk to the eye and corneal graft, and may be performed as a secondary procedure. Posterior chamber IOL implants, either iris fixated or scleral fixated, may be better alternatives over an anterior chamber IOL implants since the latter may result in graft failure,9,10 though this debate has not yet been fully resolved.
In all of our patients, the trauma occurred within 2 years after keratoplasty although they were followed up for up to 7.5 years after surgery. Most ocular trauma (11 patients [79%]) occurred during the first postoperative year. Similar findings were reported in other studies.2,3 In only a few cases, the injury occurred more than 1 year and up to 13 years following keratoplasty.4 During the first postoperative year, ocular contusive trauma caused wound dehiscence even following a mild injury such as a trivial eye poke, especially when no vascularized scar tissue in the donor-host interface was apparent. Over time, surgical wound rupture was usually the consequence of more severe injury.
Recommendations regarding the postoperative period for the use of protective eyewear have not yet been reported.1- 5 Recommendation for constant lifetime use of protective eyewear may be inconvenient for some patients even though it is mandatory after penetrating keratoplasty. Since most of traumatic wound dehiscences occurred during the first postoperative year, rational, more balanced recommendations may be provided to the patients who receive corneal transplants. The use of protective eyewear during the daytime and eyeshields during sleep should be absolutely required at least during the first postoperative year and preferably for one's lifetime. Our experience is somewhat disappointing since we found that patients frequently fail to follow these instructions. Only 2 of the 14 patients in our series were wearing protective eyewear at the time of trauma. Further studies should be designed to assess the rate of increase in the tensile strength of the corneal graft wound with time, to define more precisely the recommended period for the protective eyewear in the patient who has undergone corneal transplantation.
Accepted for publication May 26, 1998.
Presented in part at the Association for Research in Vision and Ophthalmology Annual Meeting, Fort Lauderdale, Fla, May 15, 1997.
Corresponding author: Shimon Rumelt, MD, Department of Ophthalmology, Western Galilee–Nahariya Medical Center, PO Box 21, 22100 Nahariya, Israel.