Three prospective, randomized, parallel-group, open-label, sham-controlled trials were conducted to evaluate the safety and effectiveness of riboflavin ophthalmic solution/UVA irradiation for performing corneal collagen cross-linking. These trials were sham-controlled for the first 3 months and had a total duration of 12 months for safety and efficacy evaluations. Study 1 enrolled 58 patients with progressive keratoconus and 49 patients with corneal ectasia following refractive surgery. Study 2 enrolled 147 patients with progressive keratoconus, and Study 3 enrolled 130 patients with corneal ectasia following refractive surgery. In each study, patients had one eye designated as the study eye and were randomized to receive one of two study treatments (CXL or sham) in their study eye at the baseline visit. The patients were evaluated at Day 1, Week 1, and Months 1, 3, 6, and 12. At Month 3 or later, patients had the option of receiving CXL treatment in both the sham study eyes and non-study eyes and were followed-up for 12 months from the time of receiving CXL treatment.
Approximately 56% and 89% of the sham study eyes in patients with progressive keratoconus received CXL treatment by Month 3 and Month 6, respectively. The average age of keratoconus patients was 33 years. The average baseline Kmax value was 61 diopters. For corneal ectasia patients in Study 1 and Study 3, approximately 60% and 90% of the sham study eyes received CXL treatment by Month 3 and Month 6, respectively. The average age of corneal ectasia patients was 43 years and the average baseline Kmax was 55 diopters. A majority (93%) of the corneal ectasia patients had LASIK only, 5 (3%) patients had photorefractive keratectomy (PRK) only, and 8 (4%) patients had both LASIK and PRK.
In each study, the maximum corneal curvature (Kmax) was assessed at baseline, Months 1, 3, and 12. The CXL-treated eyes showed increasing improvement in Kmax from Month 3 through Month 12 (Figure 1). Progressive keratoconus patients had an average Kmax reduction of 1.4 diopters in Study 1 and 1.7 diopters in Study 2 at Month 12 in the CXL-treated eyes while the sham eyes had an average increase of 0.5 diopter in Study 1 and 0.6 diopter in Study 2 at Month 12; the difference (95% CI) between the CXL and sham groups in the mean change from baseline Kmax were -1.9 (-3.4, -0.3) diopters in Study 1 and -2.3 (-3.5, -1.0) diopters in Study 2.
For corneal ectasia patients, at Month 12, the CXL-treated eyes had an average Kmax reduction of 1.0 diopter in Study 1 and 0.5 diopter in Study 3 while the sham eyes had an average increase of 1.0 diopter in Study 1 and 0.5 diopter in Study 3; the treatment difference between the CXL and sham groups was: -2.0 (-3.0, -1.1) diopters in Study 1 and -1.1 (-1.9, -0.3) diopters in Study 3.
Post-baseline missing data were imputed using last available Kmax value. For the sham study eyes that received CXL treatment after baseline, the last Kmax measurement recorded prior to receiving CXL treatment was used in the analysis for later time points. In Study 3, four patients in the CXL group had missing baseline Kmax value and were excluded from the analysis.