According to this study, transepithelial CXL (epi-on) is less effective in decreasing maximum keratometry but is superior in corrected distance visual acuity (CDVA) outcomes and similarly safe. The conclusions drawn by the authors do not change our opinion or current clinical practice, which is epi-off in all standard cases, for the following reasons:
The authors performed a meta-analysis but based on only three trials. They also claim that transepithelial CXL has reported better visual outcomes. Before any conclusions can be drawn, it is important to have a closer look at the data.
The study by Bikbova showed superior CDVA for epi-on only at 12 months, whereas at 24 months, CDVA was comparable in both epi-on and epi-off groups. The study by Soeters specifically states that, “When analyzing the data without the 1-month results, there is no significant difference between the 2 groups.”
Furthermore, the primary goal of CXL is to arrest progression of the disease and not necessarily improving visual acuity. Although the intention to put emphasis on visual acuity is understood, the data obtained from the three studies cannot be compared in a sound manner because the CXL protocols used in the three studies differ substantially. Bikbova and colleagues used the Dresden protocol for both epi-off and epi-on with 10 minutes of iontophoresis. Lombardo and colleagues used the Dresden protocol for the epi-off group but used a different irradiation/time setting of 9 minutes at 10 mW/cm2 for the epi-on group with iontophoresis. Finally, Soeters used the Dresden protocol for both epi-off and epi-on but did not perform iontophoresis in the epi-on group.
This discrepancy is of particular concern because recent research from our group has shown that even changing the irradiation/time parameters while keeping the mode of administration identical (Torres and colleagues, ARVO 2018, accepted) in iontophoresis-mediated epi-on CXL may lead to different results due to the central role of oxygen diffusion. In summary, no conclusion can be drawn using three different studies with three entirely different irradiation/time and riboflavin administration modalities.
In the future, a large randomized controlled trial would provide better evidence and draw more accurate conclusions because the clinical trial protocol (eg, materials, methods, inclusion/exclusion criteria, follow-up, etc.) would be more consistent. LeLorier and co-authors highlighted in a paper published some years ago in the New England Journal of Medicine how often and why there are “discrepancies between meta-analyses and subsequent large randomized, controlled trials.”
LeLorier J, et al. N Engl J. Med. 1997;doi:10.1056/NEJM199708213370806.
Farhad Hafezi, MD, PhD and Nikki Hafezi, MAS IP ETHZ
ELZA Institute AG