Oxygen-Enriched Epithelium-On CXL Meets Trial Endpoint

Key Takeaways

• Month 12 Kmax findings favored treatment enough to meet the prespecified clinical success threshold.
• Month 6 analyses and sensitivity analyses supported the same directional efficacy pattern.
• No serious ocular adverse events or severe treatment-related study-eye adverse events were reported, and punctate keratitis was the most common ocular adverse event.

A phase 3 randomized comparison in 312 eyes from 208 patients found that oxygen-enriched epithelium-on corneal cross-linking outperformed a sham/placebo procedure in keratoconus. The trial assigned 200 eyes to treatment and 112 to the comparator, and the prespecified month 12 Kmax endpoint met the study's predefined criteria for success.

The phase 3 multicenter trial ran at 28 U.S. sites from January 2023 through June 2024 and used 2:1 randomization to treatment or sham/placebo. Patients and the sponsor were masked, visual-outcome assessors were masked from month 1, and the treating investigator was not masked. Eligibility was listed as 12 to 55 years in the methods, while enrolled participants were 13 to 51 years old across 312 randomized eyes from 208 patients in the abstract. With intact epithelium, investigators used topical anesthetic, riboflavin 0.239% then 0.177%, and humidified oxygen to about 90% or greater surface saturation. UV-A was delivered at 365 nm and 30 mW/cm2 for 11 minutes 6 seconds in a 1-second on and 1-second off cycle, totaling 10 J/cm2. The primary endpoint was the between-group least-squares mean change from baseline in Kmax at month 12, with success defined by p<0.05 and at least a 1.0-D mean difference.

At month 12, Kmax improved by 0.5 D in the treatment group and worsened by 0.4 D in the sham/placebo group. The between-group least-squares mean difference was -1.0 D, with a 95% CI of -1.3 to -0.6 and p<0.0001. At month 6, Kmax improved by 0.4 D with treatment and increased by 0.1 D with sham/placebo, for a between-group difference of -0.6 D (95% CI, -0.9 to -0.2; p=0.0008). Sensitivity analyses were consistent with the primary result, and both the primary time point and the earlier analysis favored treatment.

No serious ocular adverse events or severe treatment-related adverse events in the study eye were reported through month 12. Ocular adverse events occurred in 61 of 200 treated eyes, or 30.5%, and in 12 of 112 sham/placebo eyes, or 10.7%, with most treatment-group events reported by month 3. Punctate keratitis was the most common ocular adverse event, affecting 13 of 200 treated eyes and 2 of 112 sham/placebo eyes. Non-ocular adverse events occurred in 11 of 164 treated patients and 2 of 44 sham/placebo patients, with no severe or treatment-related non-ocular events. Corneal thickness, endothelial cell count, intraocular pressure, and central subfield thickness showed no clinically meaningful change over 12 months.

Subgroup analyses were presented by age, sex, race, and baseline keratoconus severity, with significant between-group differences in patients younger than 30 years and in male, female, white, and mild-disease subgroups. Results did not reach significance in patients aged 30 years or older, in non-white participants, or in moderate to severe disease subgroups, which the authors linked to small samples.