Young patients with keratoconus face two problems: disease progression and corneal shape irregularity leading to poor vision even in spectacles.

Corneal collagen cross-linking (CXL) is a new treatment designed to halt disease progression in keratoconus. The aim is to stiffen the cornea thereby preventing further shape deterioration.

Topography or wavefront guided transepithelial photorefractive keratectomy (transPRK) uses the excimer laser (the laser used to correct sight in 'laser eye surgery') to reduce corneal shape irregularity in early stage keratoconus, reducing dependence on contact lenses.

In transPRK, the corneal skin layer is removed in a well controlled, no touch procedure, preparing the cornea for CXL. Performing both treatments simultaneously (combining both procedures in one operation) may offer several advantages over performing CXL first then waiting for corneal shape to stabilise before performing transPRK. In particular, visual rehabilitation may be faster. This study aims to evaluate visual recovery after simultaneous CXL and transPRK in progressive early stage keratoconus. Visual recovery in these patients will be compared with results for a similar group of patients with early stage keratoconus who have already been treated with CXL alone.

Features which distinguish this trial from previous trials of combined photorefractive keratectomy (PRK) and corneal collagen cross-linking (CXL) are: a rapid, pulsed light, CXL protocol; and a treatment programming algorithm for PRK designed to target higher order aberrations only with no compensatory additional laser corneal tissue removal.

Excimer laser treatment will be performed with the Schwind Amaris 750S laser (www.eye-tech-solutions.com). Unique features of this system utilised here include:

• Pre-programmed transepithelial ablation - laser removal of the minimum area of corneal epithelium required for PRK promoting more rapid recovery than conventional methods (20% alcohol application and manual epithelial removal). An 8mm diameter ablation zone will be applied throughout.
• Enhanced algorithms for minimal laser tissue removal - in this trial, the investigators are targeting higher order aberrations only (coma in particular) with the aim of improving spectacle corrected visual acuity without regard to the predicted spherocylindrical outcome. Unlike other current excimer laser platforms, Schwind Amaris treatment programming software allows treatment for irregular astigmatism without additional compensatory laser tissue removal to correct the spherocylindrical change induced by treatment of higher order aberrations. Higher order aberrations can also be treated selectively. The investigators will use an algorithm step that targets only aberrations (up to 6th order Zernike polynomials) with a value greater than 2 standard deviations from the population mean in normative data. These treatment planning steps allow the laser to create a large diameter treatment with minimal treatment depth.

Laser epithelial removal alone (transPTK at depth 55µm) removes a maximum 65µm of tissue in the corneal periphery. For patients with 390µm at the thinnest point, transPTK will therefore leave above 325µm residual stromal thickness prior to CXL. This is in line with recommendations for minimum stromal thickness after epithelial removal in the CXL protocol used here. Limited stromal reshaping is achieved in this simple embodiment of transPRK for keratoconus by taking advantage of the masking effect of the corneal epithelium, which tends to be thinnest over the cone apex. Where the thinnest point is greater than 390µm, further reductions in corneal shape irregularity can be produced by adding either wavefront or topography guided additional stromal ablation using custom programming on the Schwind Amaris laser. Ocular wavefront (aberrometry) guided smoothing will be used for patients with a 5.5mm or larger pupil at scanning and a consistent scan sequence (3 scans within 0.5 dioptre (D) spherical equivalent refraction). Corneal wavefront (topography) guided smoothing will be used for patients with ocular wavefront scans which do not meet these criteria. In all cases, a minimum corneal stromal thickness prior to CXL of 325µm will be preserved.

Immediately after PRK, corneal collagen cross-linking will be performed using the following protocol

Riboflavin soak: 10 minutes total soak time; application of 0.1% riboflavin preparation (VibeX Rapid - www.avedro.com) each 2 minutes with gentle balanced salt solution irrigation to remove excess riboflavin prior to UV light exposure.

UV light exposure: Total treatment time 8 minutes (370nm wavelength; 30mW/cm2 irradiance; 4 minutes total UV exposure time, pulsed 1.5 seconds on 1.5 seconds off; Avedro KXL I light source)

Mitomycin C will not be used. A bandage contact lens will be applied at the end of treatment and a standard post photorefractive keratectomy topical and systemic drug treatment regimen will be used in every case with initial clinical review 1 week after surgery.

Inclusion Criteria:

• Patients with progressive stage II or III keratoconus
• CDVA < 0.00 logMAR or subjective problems with spectacle corrected visual quality (ghost images or light scatter symptoms)

Exclusion Criteria:

• Active ocular surface disease
• Minimum corneal thickness <390µm (leaving 325µm residual stromal thickness after transPTK - in line with minimum thickness recommendations for the study CXL protocol)
• Vulnerable groups (patients whose capacity for giving informed consent to participate in the trial may be impaired)

Matched historical control cases of CXL treated using the same CXL protocol as the TransPRKCXL group prior to commencement of recruitment. These control cases were consecutive cases with 2 year follow up data, matched for age, corneal pachymetry, and corneal curvature with the TransPRK group

2 years following the intervention