We really appreciate the comments by Drs Kymionis and Diakonis regarding our recently published work on 4-year follow-up of corneal confocal microscopy analysis after photorefractive keratectomy (PRK) plus cross-linking (CXL) procedure in keratoconic corneas, and we would like to thank the authors for allowing us to clarify some issues emerging in this article.

Our work was focused exclusively on microstructural corneal changes after the PRK plus CXL procedure, but the initial study group included 34 eyes of 17 keratoconic patients, assigned to 2 groups of treatment: the worse eye underwent PRK plus CXL (17 eyes), and the better eye (fellow eye) underwent CXL alone (17 eyes). Here, we would like to comment on our results in the CXL group, as Drs Kymionis and Diakonis suggest.

Regeneration of corneal epithelium was detected by 10 days after CXL at the latest. Anterior stromal keratocyte nuclei were present in no eyes (n = 0) at 1 postoperative month; in 88.2% of eyes (n = 15) at 3 and 6 months; and in 100% of eyes (n = 17) at 12 months. Mid-stromal keratocyte nuclei were present in no eyes (n = 0) at 1 and 3 postoperative months; in 23.5% of eyes (n = 4) at 6 months; in 58.8% of eyes (n = 10) at 12 months; and in 100% of eyes (n = 17) at 18 months. The posterior stroma and corneal endothelium were unchanged, compared with the baseline state, at each postoperative time point. Our results, in eyes treated by CXL alone, showed that keratocyte nuclei in anterior stroma were decreased at 1 and 3 months postoperatively and then tended to increase up to 12 months; in mid-corneal stroma, keratocyte nuclei were more reduced, and for a longer time, especially close to the boundary of the treated area. Our data are consistent with those in the literature. Eighteen months after CXL, when keratocyte density was restored to preoperative values throughout the corneal thickness, needle-like structures with elongated keratocytes were still recognizable at the boundary of the treated area, persisting up to 48 months.

We agree with Drs Kymionis and Diakonis’ comment about the possible role of the Bowman membrane in ultraviolet (UV) light absorption from the corneal stroma, and with the concept that its absence may facilitate UV-A absorption and cause a more intense treatment. We are unable to comment on the role of mitomycin C because we did not use it in any case.

We thank Dr Kymionis and Dr Diakonis once again for their illuminating comments and for giving us the opportunity to add some important information to what was already reported in our work.