In their recent article, Kato and Associates described the outcomes of topography-guided conductive keratoplasty (TGCK) for advanced keratoconus in 21 eyes. The authors concluded that TGCK is effective in reshaping corneal configuration in eyes with keratoconus, without serious complications, and in avoiding or delaying corneal transplantation. Their outcome measures for efficacy were changes in visual acuity, manifest refraction, and corneal curvature. The outcome measures for safety were changes in endothelial cell counts, intraocular pressure, and occurrence of corneal infection or perforation.
Regarding efficacy, the authors observed statistically significant improvements in all outcome measures at 1 week and 1 month after surgery as compared with baseline. However, we found that the 95% confidence intervals around the mean preoperative and postoperative values of visual acuity, manifest refraction, and corneal curvature were grossly overlapping at all time points (Figures 4, 6, 8, and 9). The authors’ claim that 70% of eyes showed improved or stable vision throughout the follow-up period is also unsubstantiated, because only 10 (48%) eyes completed 1-year of follow-up. The authors’ conclusion that TGCK is able to reshape the cornea seems to be based solely on the ability of 11 of the 21 eyes to tolerate rigid gas permeable contact lenses better after surgery, without any objective evidence to support this. The selection criteria that were used for this procedure are also unclear. The authors state that all 21 eyes had advanced keratoconus, whereas preoperative keratometry was less than 48 diopters in 6 eyes (Cases 7, 10, 11, 16, 17, and 19). Patients with poor vision resulting from resolved hydrops and stromal scars also were included in this study, and not surprisingly, these eyes did not improve after surgery.
The possible complications of TGCK in keratoconus could be short term (rise in intraocular pressure, corneal melt, or infection) and long term (gradual endothelial loss or ectasia). The intraocular pressure was measured in only 8 eyes at 1 week after surgery, and there is no mention of the immediate postoperative values. It is indeed surprising that corneal tomography was possible in all eyes, but applanation tonometry was not, because of stromal scars and epithelial irregularities. Similarly, endothelial counts were obtained in only 6 eyes at 1 month; the values at 6 months or 1 year after surgery have not been provided. This omission is significant because although conductive keratoplasty is safe in hyperopic eyes, its effect on the endothelium in eyes with thin corneas is unknown.
Unlike collagen cross-linking, TGCK is not intended or supposed to arrest progression of keratoconus. Therefore, patients with nonprogressive advanced keratoconus with a clear visual axis and contact lens intolerance are ideal candidates in whom the efficacy and safety of this procedure can be evaluated. In contrast, the authors chose a heterogenous group of patients with progressive or nonprogressive disease, stromal scars, and varying degrees of corneal ectasia. The ill-defined selection criteria, small sample size, short follow-up, and lack of objective evidence regarding safety and efficacy are limitations of this study that question the applicability of the study results to eyes with advanced keratoconus awaiting corneal transplantation.