Implantation of a corneal graft–keratoprosthesis for severe corneal opacity in wet blinking eyes

5 Safety

This section describes safety outcomes from the published literature that the Committee considered as part of the evidence about this procedure. For more detailed information on the evidence, see the interventional procedure overview.

5.1 Retroprosthetic membrane formation was reported in 54% (22/41) of eyes in a case series of 37 patients who had type I corneal graft–keratoprosthesis implantation at a mean follow‑up of 22 months. Fourteen patients needed treatment with yttrium‑aluminium‑garnet (YAG) laser (3 of them needed additional surgical membranectomy) and 8 patients did not need any treatment. In 6 patients, retroprosthetic membrane formation was secondary to concomitant surgical procedures. Epiretinal membrane formation was reported in 1 patient in a case series of 40 patients (42 eyes) at 1‑year follow‑up (further details were not reported).

5.2 Corneal melting was reported in 24% (10/42) of eyes in the case series of 40 patients at a mean follow‑up of 64 months. This occurred in patients with Stevens–Johnson syndrome and the melt led to further morbidity, infection and implant extrusion.

5.3 Chorioretinal adhesion problems (retinal detachment with or without choroidal detachment or retinoschisis) were reported in 27% (11/41) of eyes in the case series of 37 patients at a mean follow‑up of 22 months. Six patients needed surgery: 5 had vitrectomy and 1 had trans‑scleral cyclophotocoagulation.

5.4 Glaucoma or increased intraocular pressure was reported in 81% (34/42) of eyes (in 5 eyes of people with, and 29 eyes of people without, Stevens–Johnson syndrome) in the case series of 40 patients at a mean follow‑up of 64 months. Increased intraocular pressure was noted in 19% (8/42) of eyes, glaucoma was newly diagnosed in 30% (13/42) of eyes and 30% (13/42) of eyes with preoperative glaucoma had disease progression. All were treated with anti‑glaucoma drugs (mean 2.7 drugs per eye), but 12 needed surgical interventions at a mean follow‑up of 22.1 months, corresponding to a glaucoma surgery rate of 0.062/eye year.

5.5 Infectious endophthalmitis was reported in 16% (20/133) of patients in a case series of 150 patients (158 eyes) at 7‑year follow‑up (further details were not reported). Infectious keratitis was reported in 21% (9/42) of eyes (3 eyes of people with autoimmune disease and 6 eyes of people with non‑autoimmune disease) in the case series of 40 patients at a mean follow‑up of 64 months (further details were not reported).

5.6 Sterile corneal necrosis was reported in 20% (26/133) of patients in the case series of 150 patients (158 eyes) at 7‑year follow‑up (further details were not reported).

5.7 Sterile vitritis was reported in 4% (4/101) of eyes in an international series and 11% (10/94) of eyes in a US series in a retrospective case series of 194 patients at a mean follow‑up of 14.2 and 24.1 months (further details were not reported).

5.8 Cystoid macular oedema was reported in 10% (13/133) of patients in the case series of 150 patients (158 eyes) at 7‑year follow‑up (further details were not reported).

5.9 Persistent epithelial defects were reported in 10% (10/101) of eyes in the international series and 36% (34/94) of eyes in the US series in the retrospective case series of 194 patients at a mean follow‑up of 14.2 and 24.1 months (further details were not reported).

5.10 Vitreous haemorrhage was reported in 10% (4/42) of eyes in the case series of 40 patients at 1‑year follow‑up (further details were not reported). Choroidal haemorrhage was reported in 3% (4/122) of patients in a case series of 122 patients (126 eyes) at 6‑month follow‑up (further details were not reported).

5.11 Device leaks at the keratoprosthesis stem (through the cornea‑anterior front plate interface of the device) were reported in 3 eyes (at a mean of 13.7 months) after type I corneal graft‑keratoprosthesis implantation in a case report of 3 patients. In 1 patient, the leak was not evident; in the second patient, a repeat keratoprosthesis implantation was needed to stop the persistent leak; in the third patient the persistent leak was repaired with glue.

5.12 Traumatic wound rupture (at the graft–host junction) at an average of 4.2 months after type I corneal graft–keratoprosthesis implantation was reported in 3% (4/136) of eyes in the case series of 122 patients. In 2 eyes, the device was extruded and therapeutic penetrating keratoplasties were performed, but vision deteriorated. In 2 eyes with wound rupture, suturing of the wound was done. Vision improved in 1 eye and in the other it was stable.

5.13 Occlusive vasculopathy (peripheral occlusive vasculitis and ischaemia of the entire retina) was reported in 5% (2/41) of eyes in the case series of 37 patients at a mean follow‑up of 22 months (further details were not reported).

5.14 Corneal infiltrate was reported in 12% (12/101) of eyes in the international series and 10% (9/94) of eyes in the US series in the retrospective case series of 194 patients at a mean follow‑up of 14.2 and 24.1 months (further details were not reported).

5.15 Scleritis was reported in 1 patient in the case series of 122 patients (126 eyes) at 6‑month follow‑up (further details were not reported).

5.16 Chronic hypotony was reported in 9% (6/67) of patients in a case series of 68 patients at a median follow‑up of 18.5 months after type I corneal graft–keratoprosthesis implantation. The incidence of chronic hypotony was 3.7% at 1 year (95% confidence interval [CI] 0.9% to 14%) and 13.3% at 2 years (95% CI 5.5% to 30%). All eyes had retroprosthetic membranes and decreased visual acuity and 5 eyes had previous history of glaucoma or ocular hypertension. Four patients had pars plana vitrectomy and silicone oil injection and reported increased vision ranging from 'hand motion' to 20/400. One patient with 1 affected eye deferred treatment and the eye progressed to phthisis bulbi needing enucleation. One eye had pre‑phthisis and no surgery was needed.

5.17 Posterior capsular tear was reported in 1 patient during the surgery in the case series of 40 patients (further details were not reported).

5.18 Sterile corneal ulceration at the graft–optic junction was reported in 22% (2/9) of eyes (1 after 52 months and 1 at 10 months), in a case series of 9 patients (9 eyes) with failed interventions for chemical and thermal injury. Both devices were removed and replaced, 1 also had a concomitant retinal detachment repair. Vision deteriorated in 1 eye and the other eye developed endophthalmitis and became blind and painful, and was enucleated.

5.19 Horizontal diplopia after type I corneal graft‑keratoprosthesis implantation was reported in a patient with a history of trauma and a series of failed corneal transplants. Strabismus surgery restored binocular vision.

5.20 Pigmented deposit on the keratoprosthesis (a large central black deposit on the bandage contact lens on the front plate of the device) associated with the use of topical ibopamine as a treatment for chronic hypotony was reported in a patient implanted with a type I device. Postoperatively, vision improved to a best corrected visual acuity (BCVA) of 20/200, but after 3 months, vision deteriorated because of the pigmented deposit. This was treated by removing the bandage contact lens and changing to a daily disposable contact lens and regular cleaning of the front plate with diluted baby shampoo and surgical sponges.

5.21 In addition to safety outcomes reported in the literature, specialist advisers are asked about anecdotal adverse events (events which they have heard about) and about theoretical adverse events (events which they think might possibly occur, even if they have never done so). For this procedure, specialist advisers reported no anecdotal or theoretical adverse events.