The purpose of this study is to demonstrate the safety and efficacy of the corneal collagen crosslinking with riboflavin and ultraviolet A in aim to increase the resistance of the graft used as a support for the Boston keratoprosthesis (KPro) type I against corneal melting (keratolysis or sterile necrosis).

The Boston type 1 keratoprosthesis (KPro) is an artificial cornea that restores the clarity of the visual axis. It is indicated in patients for whom conventional corneal transplantation offers a very low probability of success. Several innovations have led to improved outcomes following implantation of a KPro. However, retention of the KPro varies between 83 and 100% in the most recent series. The extrusion of the KPro is usually caused by the melting of the corneal tissue (keratolysis or sterile necrosis) used as a support.

This corneal melting is mediated by enzymes from the class of the matrix metalloproteinases (MMP). Several different types of insults may lead to an excess of these enzymes. Chronic inflammation of the ocular surface is undoubtedly the best recognized risk factor. Indeed, autoimmune diseases such as Stevens-Johnson syndrome, Lyell syndrome (toxic epidermal necrolysis, TENS) and mucous membrane pemphigoid, has the highest rate of corneal melting post KPro. Moreover, the development of a retroprosthetic membrane has recently been recognized as a risk factor for melting. Furthermore, the dryness of the corneal epithelium, due to insufficient tear production or an alteration of the blink reflex of the eye, can also lead to an overexpression of MMP. Finally, infectious keratitis can lead to significant thinning of the corneal tissue, even once the infectious process is resolved.

Corneal collagen cross-linking is a technique approved by Health Canada for strengthening the biomechanical properties of the cornea. The crosslinked corneas become more resistant to collagenase and other MMP. The use of a crosslinked corneal graft as a support for the KPro is an interesting approach to the prevention of corneal melting.

A prospective, randomized, controlled and double-blind study will be conducted with patients receiving a Kpro at the Centre Hospitalier de l'Université de Montréal. Forty patients will be randomized into two groups, half will receive KPro in a crosslinked graft-support, while the other half will receive a usual graft-support. Patients will be met by their surgeon to discuss the risks, benefits of the KPro type I and alternatives treatments. Patients accepting KPro surgery, will be informed of the nature and course of the study and will be offered to participate in the study. The Eye bank of Canada will manage the randomization and maintain the codes identifying the crosslinked corneas from the untreated corneas, keeping both the surgeon and the patient blinded. Randomization will be done through a free application (http://www.randomizer.org/form.htm). A series of 10 numbers will be generated. Even numbers indicate crosslinked corneas and odd numbers indicate non-crosslinked corneas. The order of the digits will increase as for each subject is enrolled in the study. The procedure for corneal collagen crosslinking will be performed at the Eye Bank of Canada in a similar procedure to the treatment for keratoconus. Under sterile conditions, the corneo-scleral button will be inspected to meet the standard of care. If the patient is randomized to the crosslinked group, then the cornea will be treated with crosslinking as described further. If the patient is randomized the control group, the cornea will not be treated with crosslink but will be deepithelialized and soaked with riboflavin drops as described further.

The surgeon will receive the graft and operate according to the standard procedure. The KPro surgery and postoperative care will be performed in the standard way and thus will be the same for both study groups. Postoperative follow will be held at 1 day, 1 week, 2 weeks, 1 month and 3 months and will continue subsequently every 2-4 months depending on the judgment of the surgeon. These visits will include measurement of visual acuity and intraocular pressure and a full slit-lamp examination. In particular, the presence of corneal melting, leakage of aqueous humor, corneal infection and extrusion will be noted. Meanwhile, an imaging allowing quantification of the thickness of the cornea graft-support (optical coherence tomography anterior segment) will be performed at least once during the first three months postoperative. This imaging will be repeated at 1, 2 and 5 years.

The prevalence of various complications (melting, leak, infection, extrusion) will be compared between the two groups with the Fischer exact test. Also, the time between surgery and the occurrence of complications will be compared using the Student t test. Finally, survival analysis of Kaplan-Meier will be performed for 1) the occurrence of corneal melting and 2) maintaining a visual acuity greater than 20/200.

Inclusion Criteria:

• Candidate for KPro type I
• Capacity to give written consent
• Ability to be followed for the duration of the study

Exclusion Criteria:

• Participation in another interventional study
• Failure to wear a therapeutic contact lens due to abnormalities of the eyelids.
• Inability to give written consent

Contraindications to the KPro type I:

• Severe dryness with keratinization of the ocular surface
• Intraocular tumor
• Terminal glaucoma
• Inoperable retinal detachment
• Phthisis bulbi