Corneal cross-linking treats ectasia after refractive surgery
A multicenter study showed the procedure improves maximum keratometry value and corrected and uncorrected distance vision.
A new multicenter trial of the U.S. Crosslinking Study Group showed that corneal cross-linking is effective in improving the maximum keratometry value and corrected and uncorrected distance vision in patients with corneal ectasia after refractive surgery.
The study was performed in order to achieve FDA approval of CXL for this indication. It involved 11 sites in the U.S. and enrolled 179 patients randomly assigned to a treatment group or a sham control group. In the treatment group, riboflavin was administered followed by ultraviolet A irradiation according to the standard Dresden protocol. Patients in the control group received riboflavin without UVA exposure and received standard CXL treatment after 3 months if needed.
One hundred sixty-six eyes had previously undergone LASIK, eight eyes had undergone LASIK with PRK re-treatment, and five had undergone PRK.
“This study and its results are very important. They represent a further step forward after the recent approval of CXL for keratoconus in the U.S.,” Peter S. Hersh, MD, lead author and medical monitor of the trial, told Ocular Surgery News.
“In addition to proving efficacy, this study demonstrates the safety of this technique. Cross-linking is the first procedure that we have to decrease progression of corneal distortion in patients with ectasia,” he said.
Stabilizing the cornea
The goal of CXL in this study was to stabilize ectatic corneal processes.
“If you have a thin point in your car tire, it will probably continue to balloon out. Similarly, in the case of corneal ectasia, there is a tendency to get worse over time. The goal is to diminish progression,” Hersh said.
The maximum keratometry value on corneal topography was selected as the main outcome indicator because it is a reproducible corneal topography index and a good proxy for the steepness of the ectatic cone. Study results showed that maximum keratometry value in treated patients diminished by 0.7 D on average over the 12 months of follow-up compared with the control group in which continued progression and worsening of the maximum keratometry value were reported.
“In the treatment group, only three eyes (4%) continued to progress over the course of the year by 2 D or more,” Hersh said.
Side effects were minimal. Only one severe adverse event was reported in a patient who had epithelial ingrowth beneath the LASIK flap and needed the flap to be lifted to remove the ingrown cells.
In the CXL treatment group, there was a significant improvement of five letters (approximately one line) of corrected distance visual acuity at 12 months, while the control group lost 0.3 letters. However, while the majority of treated eyes remained stable, 32% of patients gained and three eyes lost 10 or more letters. No obvious clinical correlate was found for those patients in whom vision worsened; all three eyes had stable topography with no change in maximum keratometry value over the 12 months of follow-up.
In an attempt to identify factors that may determine the good or poor response to treatment, Hersh carried out a multifactorial analysis of patients treated with CXL in his own clinic.
“We looked at many preoperative factors such as age, gender, steepness of the cone and VA, and found that the only indicator for actual corneal flattening from cross-linking was a higher steepening at baseline. Looking at ectasia and keratoconus together, we found that patients who had cone steeper than 55 D were 5.4 times more likely to have a flattening effect from the treatment. Regarding treatment failure, we found no preoperative factors suggesting that the treatment would not stabilize the cornea,” he said.
Similarly, the only indicator for improvement of corrected vision was worse vision at baseline. Patients who were 20/40 or worse had a 5.9 times greater chance of improvement, while patients with vision better than 20/40 had a slightly increased risk of losing one line of vision with the treatment.
“These findings show that patients with progressive disease and diminishing vision are overall the best candidates,” Hersh said. “They also bring up the necessity to discuss risks and benefits with patients. Patients with progressive disease and good vision should know that they have a chance to experience a slight vision loss, usually transient, after CXL treatment.”
Screen carefully, treat early
The risk for developing ectasia after refractive surgery has diminished significantly in recent years due to improved ability to screen patients for signs of keratoconus with corneal topography techniques.
“We are now better educated on how to screen patients prior to surgery and follow them over time with topography to detect possible problems as soon as they arise,” Hersh said.
Awareness should be raised on the necessity to treat early with CXL.
“Ectasia after refractive surgery as well as keratoconus per se are not going to get better by themselves. Therefore, we need to identify these patients early and treat them with CXL because for the large majority it is an appropriate and effective treatment,” Hersh said.
Although ectasia is a risk more related to LASIK than PRK, there are good reasons to choose LASIK whenever there are the conditions to perform it safely, according to Hersh.
“Refractive surgeons should identify risk cases, and if there is a concern with thin or irregular corneas, or concerns related to making a LASIK flap, they should do PRK or avoid surgery, depending on the case,” he said.
“It would be interesting to go back on our study and analyze the preoperative data of all the patients who entered it to see if there were indicators of the risk of ectasia,” Hersh said. – by Michela Cimberle
- Hersh PS, et al. Ophthalmology. 2017;doi:10.1016/j.ophtha.2017.05.036.
- For more information:
- Peter S. Hersh, MD, can be reached at the Cornea and Laser Eye Institute – Hersh Vision Group, CLEI Center for Keratoconus, Glenpointe Centre East, 300 Frank W. Burr Blvd., Teaneck, NJ 07666; email: firstname.lastname@example.org.
Disclosure: Hersh reports he was medical monitor for the Avedro multicenter clinical trials and was previously a consultant for Avedro.