2 The procedure

2.1 Indications and current treatments

2.1.1 The corneal endothelium is a single layer of cells comprising the cornea's innermost layer. It helps maintain corneal transparency by removing excess fluid. Endothelial dysfunction leads to corneal clouding, resulting in visual impairment. Common causes of corneal endothelial dysfunction are Fuchs' dystrophy (a genetic disorder) and degeneration (bullous keratopathy). Other causes are trauma, infection and iatrogenic damage.

2.1.2 Current surgical treatment for corneal endothelial disease is penetrating keratoplasty (PK) (whole cornea transplantation), which requires multiple sutures to anchor the donor cornea in the recipient eye.

2.2 Outline of the procedure

2.2.1 Corneal endothelial transplantation uses a range of techniques to replace diseased corneal endothelium with a cadaveric donor endothelial graft, while retaining healthy portions of the patient's cornea. It may be performed under local or general anaesthesia. A scleral incision is made and an anterior chamber tunnel created. The diseased endothelium is dissected from the cornea and donor endothelial graft inserted through the scleral incision and placed against the posterior aspect of the host cornea.

2.2.2 Topical and/or systemic antibiotics, steroids and immunosuppressants are often prescribed after surgery.

Sections 2.3 and 2.4 describe efficacy and safety outcomes which were available in the published literature and which the Committee considered as part of the evidence about this procedure. For more detailed information on the evidence, see the overview.

2.3 Efficacy

2.3.1 A randomised controlled trial (RCT) of 28 eyes reported significant improvement in mean uncorrected visual acuity (UCVA) in 13 EK-treated eyes, from 0.81 (standard deviation [SD] 0.19) to 0.60 (SD 0.20) at 6-month follow-up (p = 0.01). In 15 PK-treated eyes, no significant UCVA improvement from baseline was reported at 6-month follow-up (0.94 [SD 0.38] to 0.87 [SD 0.30]) (significance not stated). A non-randomised comparative study of 177 eyes (129 EK-treated) reported significantly better UCVA after EK than after PK at 15-month follow-up (p = 0.05), without significant difference in contrast sensitivity between the groups (median follow-up 12 months).

2.3.2 The non-randomised comparative study of 177 eyes reported that the astigmatism incidence was significantly lower after EK than after PK (p < 0.0001) (mean follow-up 15 months).

2.3.3 A registry reported significantly different 1-year graft survival rates of 77% (95% confidence interval [CI] 63 to 86) and 98% (95% CI 91 to 99) among 75 EK- and 88 PK-treated patients with Fuchs' dystrophy, respectively (p = 0.0002). Significantly different 1-year graft survival rates of 79% (95% CI 65 to 88) and 88% (95% CI 75 to 94) were reported among 55 and 76 EK- and PK-treated patients with pseudophakic bullous keratopathy, respectively (p = 0.04).

2.3.4 Specialist Advisers listed key efficacy outcomes as UCVA, visual rehabilitation speed, and quality of life measures such as the Visual Function Index (VF-14) score.

2.4 Safety

2.4.1 Two case series and an RCT of 100, 118 and 28 eyes (13 EK-treated) reported PK conversion in 2% (2/100), 9% (11/118) and 19% (3/16) of procedures planned as EK.

2.4.2 The studies of 100, 118 and 28 eyes reported need for repeat EK in 2% (2/98), 8% (10/118) and 8% (1/13) of eyes.

2.4.3 A non-randomised comparative study of 907 eyes (199 EK-treated) reported significantly lower graft rejection rates for EK (8% [15/199]) than for PK (13% [92/708]) at 2-year follow-up (p = 0.04). The non-randomised comparative study of 907 eyes also reported that graft failure after rejection was lower for EK (7% [1/15]) than for PK (28% [26/92]) (p = 0.063). A non-randomised comparative study of 177 eyes (129 EK-treated) reported no significant difference between EK and PK in graft rejection (p = 0.78) or primary graft failure (p = 0.91) rates (mean follow-up 15 months).

2.4.4 The non-randomised comparative study of 177 eyes reported that graft dislocation was significantly more common after EK than after PK (p = 0.0004).

2.4.5 In a case series of 263 EK-treated eyes, cumulative endothelial cell loss (not otherwise defined) in a subset of 34 eyes with 2-year follow-up was 34% at 6 months, 36% at 12 months, and 41% at 24 months.

2.4.6 A case series of 118 EK-treated eyes (41 with concomitant cataract surgery) reported retinal detachment in 4% (5/118) of patients (sequelae and follow-up not described).

2.4.7 The Specialist Advisers listed adverse events reported in the literature or anecdotally as graft dislocation, graft failure and rejection, interface opacification, and loss of BSCVA (best spectacle-corrected visual acuity).

2.5 Other comments

2.5.1 The Committee noted that UK Transplant Register data show lower graft survival rates after EK than after PK. The difference in graft survival between the two procedures is narrowing with increased experience in EK use. Endothelial keratoplasty can be repeated, while PK revision is more difficult. The Committee therefore felt that the current evidence on efficacy of the procedure was adequate, provided that thorough data collection continues, to allow future review of outcomes.

2.5.2 The Committee noted that the techniques for this procedure continue to evolve.

2.5.3 Anterior eye procedures are classified as having a medium risk of Creutzfeldt-Jakob disease (CJD) transmission.

  • National Institute for Health and Care Excellence (NICE)