We examined the efficacy and preoperative characteristics that affect outcomes of accelerated (9 mW/cm² for 10 minutes) corneal cross-linking (CXL).

Prospective single-center observational cohort study.

We enrolled 612 eyes of 391 subjects with progressive keratoconus (n = 589), pellucid marginal degeneration (n = 11), and laser in situ keratomileusis–induced ectasia (n = 12). We evaluated best spectacle-corrected visual acuity (BSCVA), topography, refraction, endothelial cell density, corneal thickness, haze, intraocular pressure, and visual function before and 12 months after the CXL procedure. We tabulated the proportion of those with progression of maximum keratometry (Kmax). We included participant's race, age, sex, and the presence of preoperative apical scarring and environmental allergies in a multivariable linear regression model to determine the effect of these characteristics on outcomes.

At 1 year there was no significant change in mean Kmax (n = 569). Progression of Kmax was higher in subgroups with a baseline Kmax >58 diopters (n = 191) and those 14-18 years of age (n = 53). Preoperative BSCVA, Kmax, refraction, corneal cylinder, coma, central corneal thickness, and vision function were statistically and clinically significant predictors of outcomes (P < .001). Preoperative apical scarring led to worsening haze (P = .0001), more astigmatism (P = .002), more central corneal thinning (P = .002), and was protective to the endothelium (P = .008). Race, age, and sex affected some outcomes.

Mean Kmax was stable at 1 year after accelerated CXL. Younger patients and those with a higher preoperative Kmax need to be monitored closely for progression. Preoperative BSCVA, topography, refraction, CCT, and apical scarring were significant predictors of outcomes.