We read with great interest the article by Huang and associates reporting the clinical outcomes of deep anterior lamellar keratoplasty (DALK) using the big-bubble technique in various corneal diseases. In this prospective interventional study with 115 patients (131 eyes), the researchers demonstrated that advanced keratoconus had the highest incidence of exposed Descemet membrane, whereas bacterial keratitis and moderate keratoconus showed the lowest incidence of membrane exposure. Furthermore, additional intrastromal manipulation during the surgeries may have a deleterious effect on the corneal endothelium. As much as possible, corneal specialists should avoid multiple manipulations even within the intrastromal space. Although these findings are useful for eye specialists handling such cases, we believe that several important concerns were not addressed.

First, the authors leave the readers with the impression that postoperative recovery of visual acuity after DALK was slower in cases with keratoconus than in those with herpes simplex keratitis and corneal dystrophy. They speculated that the corneal sutures and refraction shifts after surgery may account for this strange phenomenon. It should be noted that their opinion that patients with corneal dystrophy demonstrated less fluctuation in refraction than those with keratoconus after keratoplasty is based on samples from 2 different studies, which obviously lack comparability. In fact, they could have analyzed the postoperative refraction status of these groups, particularly the degree of refractive error and astigmatism. If the data were available, the authors could have evaluated the difference in the refraction status between the keratoconus group and the other groups at follow-up. In our opinion, it is best to determine the frequency and extent to which DALK improves visual acuity in these groups.

Second, in Figure 2, the authors highlight that patients in whom deep stromal air injection was attempted more frequently showed significantly lower endothelial cell density (ECD) than patients in whom the first attempt was successful. Because patients with bacterial keratitis and moderate keratoconus had the lowest incidence of Descemet membrane exposure, we wonder whether the air injection attempts varied across the different diagnoses. If they did, the authors may ignore excluding the effect of the different original diagnoses on the changes of ECD. It is agreed that the ECD was higher in the DALK groups in all studies at study completion, and in general, the differences in the ECD between the DALK and penetrating keratoplasty groups were significant at all time points at 6 months or longer after the surgery. However, whether the accelerated long-term loss of ECD in DALK is related to various original diagnoses or other causes has not yet been determined. Sufficient evidence remains to be gathered to address this issue.

Finally, as described in a recent report, we recommend that systemic and topical prophylactic antiviral agents be administered for active or quiescent herpetic stromal keratitis before surgery to prevent any herpes recurrence possibly caused by DALK.