We have read the article by Caporossi and associates with great interest. The encouraging long-term outcomes of collagen cross-linking and high index of safety that the authors have reported, is commendable. However, we believe there are a few gray zones in the article that could easily be clarified.

The authors have included patients with keratoconus progression documented “instrumentally” and clinically. Since no clear-cut guidelines to define progression exist, it would be interesting to note the criteria used by the authors to define progression in their study. Also, since initially only one eye was chosen to undergo cross-linking, it would be nice to know the criteria for selection.

The article mentions that patients included in this series were enrolled during the first 6 months of the study period and completed a minimum of 48 months of follow-up. It would be of interest to know if these were consecutive patients and if any patients were excluded, as this would affect the denominator in calculating incidence values.

The significant underestimation of pachymetry values obtained on the Orbscan (Bausch & Lomb, Inc., Rochester, New York, USA) does emphasize the fact that postoperative pachymetry values obtained on the Orbscan may be fallacious. However, it would be inappropriate to compare the values of minimal thickness of the cornea obtained on the Orbscan with those of the central cornea using ultrasound or HRT II (Heidelberg Retinal Tomography; Heidelberg Engineering, Rostock, Germany) confocal microscopy, since the 2 points may not coincide.

The subset of patients undergoing preoperative and postoperative anterior segment optical coherence tomography (OCT) measurements (Figure 1) is not mentioned. This assumes significance, as the Visante OCT was apparently available to the authors only during the latter part of the study.

The authors encountered an insignificant mean endothelial cell loss (2%). Though the authors mention this is consistent with physiological levels, literature suggests physiological cell loss is lower at 0.6% per year. The hyperopic shift and reduction in the cylindrical refractive error along with improvement in both the uncorrected and the best-corrected visual acuity are heartening to note. However, when the mean uncorrected visual acuity and the time-matched refractive error are compared, the values do not seem to corroborate, and it would be good to have pupil size measurements mentioned too.

The abstract mentions that 65% of contralateral eyes showed progression 24 months after commencement of the trial. Though these patients apparently underwent cross-linking (Figure 4), the text does not give any details about the same. We are also curious to know how the authors masked the results or the tests.

Performing cross-linking in young patients may reduce the burden of keratoplasty in the economically productive age group, but the authors’ interpretation that it is most indicated in such cases is extrapolation beyond the data, as age-specific comparative evaluation of results is not mentioned. Also, since the authors have not performed cross-linking in patients with stationary keratoconus, concluding that the procedure may improve visual acuity in such patients again is extrapolation beyond the data.