The problems of penetrating the epithelium in corneal crosslinking (CXL) may be solved, R Doyle Stulting MD, PhD told the 2016 ASCRS•ASOA Symposium & Congress in New Orleans, USA.
In his Binkhorst Medal Lecture, Dr Stulting reviewed early 12-month and 24-month data from his study under a CXL protocol that leaves the epithelium intact. This new epi-on approach achieves crosslinking results that appear to be similar to those obtained with the classic epi-off Dresden protocol out to two years – without the risks of infection, non-healing epithelial defects, corneal haze, sterile infiltrates, and perforation associated with epithelial removal.
Previous epi-on CXL attempts have not been as effective as those obtained with epi-off – primarily because the epithelium blocks the penetration of riboflavin into the stroma. There are other significant aspects of the CXLUSA protocol that may also explain the results he obtained. Previously, investigators of epi-on CXL have reported loss of effect after one year. This is the first report of improved visual acuity after epi-on CXL that remains stable from year one to year two after treatment.
The new CXLUSA epi-on protocol uses a reformulated riboflavin compound that better penetrates the epithelium, explained Dr Stulting, Director of the Stulting Research Center at Woolfson Eye Institute, Atlanta, USA. The result is corneal riboflavin concentration more than four times that obtained with previous epi-on formulations. These concentrations are similar to those obtained after epithelial removal.
LIGHT EXPOSURES
In his clinical study, riboflavin concentrations were verified by slit-lamp examination and comparison to standard photographs before patients were exposed to ultraviolet (UV) light to ensure adequate stromal saturation. No additional drops were added during UV exposure, to avoid blocking UV radiation. The UV source was pulsed to allow oxygen to diffuse into the corneal stroma between light exposures.
Of 137 keratoconus and ectasia patients treated with the new protocol, mean best corrected visual acuity improved from 20/36 before treatment to 20/27 at 12 months, with a mean gain of 1.5 lines, Dr Stulting reported. Some patients experienced gains as great as 20/100 to 20/20 at 12 months. Thirty-three patients followed to 24 months maintained these gains. Structurally, CXL demarcation lines visualised by optical coherence tomography were deeper than those reported previously with epi-on procedures.
While CXLUSA patients did not achieve as much corneal flattening as is typically seen with epi-off CXL, they did obtain a significant decrease in higher order aberrations, Dr Stulting noted. He believes that Kmax may not be the best indicator of efficacy for CXL procedures, and that high-order aberrations may be a better objective measure of functional best spectacle-corrected vision than Kmax.
“At one and two years, the technique used by CXLUSA achieves visual results that appear to be at least as good, and possibly better, than some epi-off procedures,” said Dr Stulting.
“Even if epi-on CXL does not leave corneas as resistant to progression of ectatic diseases as epi-off does, it is certainly safer and may be 'good enough',” Dr Stulting said. “If it isn’t, epi-on CXL can safely be repeated or followed by epi-off treatment,” he added.
Dr Stulting also reviewed the Ectasia Risk Score System (ERSS) he developed with colleagues at Emory University. In 2003, initial analysis of 10 eyes in seven post-LASIK ectasia patients identified thinner pre-op corneas, higher pre-op refractive error, thinner residual stromal bed and form fruste keratoconus as statistically significant risk factors (Randleman et al, Ophthalmology 2003). In 2006, analysis of eight ectasia cases without any of these characteristics identified age as another risk factor (Klein et al, Cornea 2006).
These five risk factors were validated at high statistical significance in large retrospective case control studies, and formed the basis of the ERSS in 2008 (Randleman et al, Ophthalmology 2008). Applied to the initial study population operated in 1994 and 1995, ERSS correctly identified more than 90 per cent of ectasia cases with nine per cent false negatives and four per cent false positives, Dr Stulting said. Similar results obtained by scoring a second, independent population validated the system (Randleman et al, AJO 2008).
Yet studies applying ERSS to more recently operated cases found ERSS less sensitive. Only a little over half of high-risk patients were correctly identified in populations operated after 2000 (Spadea et al, Clinical Ophthalmology 2012; Colin et al, Clin Exp Ophthalmol 2010). And new risk factors, such as anterior horizontal coma, were identified (Buhren et al, JRS 2013).
However, this is not evidence for the failure of ERSS, but a testament to its success, Dr Stulting said. The system, along with its supporting research, alerted surgeons to the identified risks, and consciously or not they avoided treating high-risk patients, he explained.
Indeed, patients with pre-op form fruste keratoconus fell from 88 per cent of ectasia cases reported in 2003 to zero in 2013. Pre-op cornea and mean residual stromal bed thickness also increased. Yet the mean age of ectasia patients fell, increasing its significance, Dr Stulting noted. “Age follows an opposite pattern because surgeons are still operating on young patients, and elimination of other factors makes age a stronger predictor of ectasia,” he explained.
Likewise, previously unidentified risk factors emerge as patients with known factors are eliminated, Dr Stulting observed. Therefore, he recommends a two-step screening process that accounts for newly identified factors without ignoring those discovered earlier. “New analysis will not identify risk factors that currently prevent patients from being treated. But future screening methodologies must recognise all risk factors lest we forget what we have learned,” Dr Stulting added.
R Doyle Stulting: dstulting@woolfsoneye.com
