Factors help identify high risk for ectasia after LASIK

Key Points

New Orleans-The primary risk factors for post-LASIK corneal ectasia are topographic abnormalities, low residual stromal bed, young patient age, low preoperative corneal thickness, and extreme myopia, according to the results of a recent retrospective study.

A weighted risk scale based on these factors should significantly improve the preoperative identification of high-risk eyes, said J. Bradley Randleman, MD, during a refractive surgery free papers session at the annual meeting of the American Academy of Ophthalmology. Dr. Randleman is assistant professor of ophthalmology, section of cornea, external disease, and refractive surgery, Emory Eye Center, Atlanta.

Preoperative risk factors that have been reported in the past include high myopia, low residual stromal bed, and forme fruste keratoconus. Recently, a position paper suggested that clinicians also watch for asymmetric inferior corneal steepening, asymmetric bowtie patterns, and skewed radial axes, Dr. Randleman said.

Dr. Randleman and colleagues performed a retrospective review of all ectasia cases in patients seeking treatment at the Emory Eye Center from 1998 to 2005 as well as those found in the English language literature through December 2005, identifying a total of 171 cases that were compared with 186 controls. They analyzed patient age, gender, preoperative manifest refraction spherical equivalent, preoperative corneal thickness, predicted residual stromal bed thickness, and topographic patterns. The topographic patterns were broken down into three categories: normal, suspected (mildly or significantly), and abnormal.

Approximately 50% of these patients presented with ectasia within 1 year of having LASIK and 80% within 2 years; however, it took up to 4 years to capture all of the cases included in the study.

Several significant differences between the ectasia cases and the controls were found. Patients who developed ectasia were younger, more myopic, had thinner corneas preoperatively, and had a significantly higher percentage of abnormal topographies and significantly thinner residual stromal bed thicknesses.

Not all of these variables were available for every ectasia case, given the retrospective nature of the study, however. To refine their analysis, the investigators defined a subgroup of 86 ectasia cases for which complete information was available and compared them with 133 contemporaneous controls. The differences between the cases and controls in this subgroup were similar to those in the full study group with the exception of preoperative refraction, which was relatively equal between the two groups, Dr. Randleman said.

In a multivariate logistic regression analysis of the subgroup, the most significant risk factors for ectasia were, in order of importance, abnormal topography, residual stromal bed, age, and preoperative corneal thickness.

This finding then was used to develop a risk score modeling system in which points were applied in a weighted fashion to any suspected or abnormal topography, any residual bed thickness less than 300 µm, any patient aged younger than 30 years, a corneal thickness less than 510 µm, or extremely high myopia. The cumulative score for each patient described the risk of ectasia.

If the score was 0 to 2, the patient was at low risk and it was considered safe to proceed with LASIK. At a score of 3, risk was considered moderate, and surgeons would be advised to consider other options and weigh variables such as refractive stability, family history of keratoconus, or a history of eye rubbing that might make the procedure more risky. However, after these variables and other options were considered, LASIK surgery could proceed with caution at the discretion of the physician. Patients with a score of 4 or higher were judged to be at high risk, and LASIK surgery would be contraindicated.

Dr. Randleman and colleagues applied this scale to their study population and found that the scores closely matched the status of the treated patients and controls.

"We found good sensitivity and specificity, with 98% of the control population being identified as low risk and 93% of the cases that developed ectasia being identified as high risk. They would not have been offered surgery based on this model," Dr. Randleman said.

This model was compared with more traditional screening methods relying on cutoff measures such as forme fruste keratoconus or a residual stromal bed thickness less than 250 µm. Results showed that the scoring system developed by Dr. Randleman and his colleagues was significantly more sensitive in detecting at-risk patients. The validity of the scoring system has not been established for surface ablation procedures because there were too few cases of this type in the study.

"However, even with this scale, some eyes will still develop ectasia, thus ectasia development does not per se constitute malpractice," Dr. Randleman said.