Corneal Collagen Crosslinking: Still Too Experimental for Keratoconus?

C. Stephen Foster, MD


June 12, 2012

The Latest Strategy for Keratoconus

A great deal of interest, time, and money has been spent on ways to stop the progression of keratoconus, to reshape the cornea, and to otherwise improve the plight of patients with keratoconus without the need for corneal transplantation. These efforts have ranged from contact lens molding/reshaping strategies to insertion of ring segments into the corneal stroma to the latest experimental strategy of corneal collagen crosslinking. This latter technique was the subject of many presentations at Association for Research in Vision and Ophthalmology (ARVO) 2012, with that of Hersh and associates[1] being one of the most interesting.

Clinical Outcomes of Corneal Crosslinking

Hersh and associates[1] reported on 85 eyes with keratoconus (n = 56) or ectasia (n = 29) which underwent corneal collagen crosslinking in a prospective, randomized controlled trial. Clinical outcomes, including uncorrected visual acuity (UCVA), best corrected visual acuity (BCVA), maximum keratometry (Kmax), corneal topography indices, corneal and ocular wavefront analysis, questionnaire-assessed subjective visual function, crosslinking-associated corneal haze measured by Scheimpflug densitometry, and corneal thickness were assessed and analyzed at baseline, 1 month, 3 months, 6 months, and 1 year.

Preoperative UCVA was 20/137 (logMAR 0.83), worsening to logMAR 0.91 at 1 month, and improving until 1 year, with UCVA of 20/117 (logMAR 0.76). Preoperative BCVA was 20/45 (logMAR 0.34), worsening to logMAR 0.38 at 1 month, and then improving until 1 year, with UCVA of 20/34 (logMAR 0.24). At 1 year, improvements in night driving, reading, diplopia, glare, halo, and starbursts were significant. Preoperative Kmax was 58.5, worsening to 59.8 at 1 month, and then improving until 1 year (Kmax 56.9). Less than one third (30.6%) of eyes flattened by 2.0 diopters (D) or more, and 3.5% steepened by 2.0 D.

Of 7 corneal topography indices, there were significant improvements in index of surface variance, index of vertical asymmetry, keratoconus index, and minimum radius of curvature (P < .001). There was a general worsening of topography indices at 1 month, followed thereafter by improvement.

Analyzing corneal higher-order aberrations, total higher-order aberrations, total coma, third-order coma, and vertical coma were all significantly reduced at 1 year after corneal collagen crosslinking treatment (P < .001). For total ocular higher-order aberrations, total higher-order aberrations, total coma, third-order coma, and trefoil all decreased (P = .01).

The mean preoperative thinnest pachymetry was 440.7 µm. After corneal collagen crosslinking, the cornea thinned at 1 month (mean change -23.8 µm; P < .001) and from 1 to 3 months (mean change -7.2 µm, P = .002), followed by a recovery of the corneal thickness at 6 months (mean +20.5 µm; P < .001). At 1 year, pachymetry remained slightly reduced from baseline to 12 months (mean change -6.6 µm; P = .01).

Corneal collagen crosslinking-associated haze was 14.9±1.93 densitometry unit preoperatively, peaking at 1 month (mean 23.4±4.40, P < .001); little change was seen at 3 months (22.4±4.79, P = .06). Densitometry decreased between 3 and 6 months (19.4±4.48, P < .001) and 6 and 12 months, but it did not completely return to baseline (mean 17.0±3.82, P < .001).

The conclusions of this very well-designed and thorough study were that the clinical outcomes of corneal collagen crosslinking follow a generally definable time course with worsening at 1 month, a return to baseline at 3 months, improvement at 6 and 12 months, and stabilization beyond 1 year. Most patients were stable or had topographic improvement over the year after corneal collagen crosslinking.