Three different protocols for inducing corneal crosslinks in subjets with progressive keratoconus will be evaluated in this randomised clinical study.

Riboflavin does not penetrate the intact corneal epithelium. Corneal cross linking (CXL) is typically performed using the so-called "Dresden protocol". The Dresden protocol states 30 minutes of UVA-radiation (3mW/cm2) but a 10 minute irradiation protocol (9mW/cm2) is frequently used. Both of the protocols involve mechanical removal of the epithelium over the central 8 mm of the corneal surface. The first days after treatment therefore involves some degree of pain, often intense, and the presence of a healing epithelial defect may be associated with development of infiltrates in the cornea. A number of approaches have been evaluated in order to promote riboflavin penetration through the intact epithelium, of which iontophoresis appears most promising. Keratoconic corneas are thin at the cone location and sometimes it is difficult to maintain the safety margin of 400 microns during corneal crosslinking. Instead of using isotonic standard riboflavin, a swelling effect of the cornea can be obtained by using hypotonic riboflavin. However, the latter has been indicated as less effective in the process of inducing cross links.

Eighty-one of 81 patients of various degrees of keratoconus will be randomised to one of the following groups: 1) CXL (UVA 9mW/cm2) using isotonic riboflavin, or 2) CXL (UVA 9mW/cm2) using hypotonic riboflavin or 3) Iontophoresis with Ricrolin with following CXL (UVA 9mW/cm2).

Hypothesis:

i) CXL with hypotonic riboflavin or CXL with Ricrolin administered by iontophoresis or CXL with isotonic riboflavin is non-inferior compared to standard CXL with isotonic riboflavin.

ii) The morphological structure post-CXL in the three different groups will be similar without any significant differences.

The iontophoresis-assisted treatment arm has been interrupted due to low efficacy in halting disease progression. The results have been published.

Inclusion Criteria:

• Progress in keratoconic eye. We define progress as an increase in Kmax of 1.0 diopter in 1 year or 0.5 diopter in 6 months. This increase in Kmax will be accepted as progression if concomitant changes tomographic parameters.

Exclusion Criteria:

• Concurrent ocular infection or corneal disease other than keratoconus.
• Pregnancy.
• Treatment with Isotretinoin.